A Large Off-Grid System [Archive] - Green Power Talk Forum

Rob Beckers

4th January 2014, 15:06

We just (almost) finished installing a relatively large off-grid system, and I thought this could be of interest to others.

This is a residence that was never supposed to be off-grid; the owners purchased a large piece of land and built a house on it (close to the main road). In the process they applied for an electrical connection with Hydro-One, and were told it was going to cost between $100,000 and $150,000 to get hooked up. As it happened, their new house was about a kilometer away from the nearest electrical pole, either way down the road...

They contacted us, and we had a good talk about what it meant to be off-grid vs. grid-tied. For one, the air conditioning that was planned to be installed had to go, but more than that, it involves a somewhat different life-style when it comes to energy. The owners were on board, and I designed a system around their electrical needs. This involved making a large spreadsheet with every light bulb, appliance, and other electrical load in the house, the peak-demand for that load, and how many hours a day it would be used. This is all added up, and the result is both the peak-load that the inverters have to supply, as well as the daily/monthly energy needs in kWh that the system has to provide. In their case we worked out that a peak-load of around 5kW would cover 95% of their needs, and they should be able to live within a 300 kWh per month allowance (10 kWh per day). As off-grid systems go this is a fairly large system. Compare this with the average grid-tied electrical user in Canada that clocks in at 1,000 kWh per month, with a 200A (50kW) connection!

We do not have much wind in our area, and so the plan was to cover the energy input largely with solar PV (also because solar panels have become so cheap). They still have to produce that 300 kWh per month even in the darkest time of winter (happens to be November over here), and to make that happen we needed 8 kW of PV. With the cheap 250 Watt Canadian Solar panels we used that worked out to 32 panels, split in strings of 4 panels (in series), and 4 of those strings going into a combiner box. That means we have two MidNite Solar MNPV6 combiners (http://www.midnitesolar.com/productPhoto.php?product_ID=77&productCatName=Combiners&productCat_ID=9&sortOrder=4&act=p), each with 4 touch-safe fuse holders, holding a 15A fuse per string.

The home owners cleared some more trees in the backyard, and we installed ground mounted racking that is seasonably adjustable; 30 degrees in summer, 60 degrees in winter (it will do 45 degrees as well, but who needs it). This is made by FabRack (http://www.fabracksolar.ca/fabrack/index.html), a local business in our area. We made a gravel pad, forms, and poured a few cubic yards of concrete to hold the racks on the ground. Installing the racking itself took less than a day.

1074

The back side shows better how this goes together; it's a spiderweb of aluminum tubing with stainless steel fasteners. Don't look too closely at the wiring: We still have to clean that up and fasten it properly. With -29C outside nobody was in a hurry to get that done just then.

1075

This particular residence was going to have propane as their fuel source, and so we specified a propane generator to get through those times in winter when the sun is not shining for days on end (or when more energy is needed than the panels can provide). Unless one is independently wealthy the choices are limited when it comes to propane generators; it is pretty much only Generac that is somewhat affordable. Generac is not the greatest of brands, with a spotty reputation, in many ways a typical "consumer quality" producer of goods, but the Onans and Kohlers are several times more expensive. Generac makes a line of LP generators, in various kW output sizes, and all are used in off-grid applications. However, their warranty is void if they find out, since most of their gensets are only meant for stand-by duty (not continuous duty as it is used off-grid). Their only exception is the Generac 6kW EcoGen (http://www.generac.com/Residential/EcoGenSeries/6kW/), that one is specifically meant for off-grid and continuous duty use. Of course, in comparison to their other gensets one pays a premium for this as well, but what's an installer to do. The 6kW size was just about right for this application, though I'd have liked something slightly larger (10kW or so). It would do, and we had our generator guy Cody install one (Fox Emergency Power Solutions (http://www.foxeps.com/), he's great if you ever need a generator installed!).

1076
1077

The reason 6kW was a little 'light' for this application was because we were planning on two Magnum MS4448PAE inverters (http://magnumenergy.com/ms-pae-series-invertercharger/). Each inverter is 4.4 kW peak continuous output (with more than twice that for short durations, to start motors and compressors such as the heater blower motor, fridge compressor etc.). These are pure sine-wave inverters that will run any load beautifully. These are also 240 Volt split-phase inverters, out-of-the-box, so we could hook this up without any modification to the electrical panel in the house.

With two 4.4 kW inverters it can supply 8.8 kW tops, which is more than the generator can deliver, so conceivably there could be the situation that the generator would kick in because the batteries are running low, and it would not be able to supply the loads. That is not likely and hopefully will never happen; the load calculations showed the need for about 5 kW, too much for a single inverter to supply but two will do it with ease. When the generator takes over it also has to recharge the batteries, so one wants a little 'headroom' beyond supplying just the house loads, another reason why 6 kW is marginal.

Besides the two inverters, we also used Magnum's panel board, auto generator start, battery monitor, and of course their controller. They all work beautifully together, and IMO Magnum presents the best bang for the off-grid buck at this time!

1079

The Magnum panels are very nicely made, and come with the main DC and AC breakers. They are a pleasure to install (certainly compared to doing all this with individual components). The only items we added were the two 80A DC breakers for the two charge controllers. All the other breakers were already present and partially wired up.

1080

To store the energy we needed batteries. A whole lot of batteries!
The norm for a regular off-grid residence is to have 3 days of 'autonomy', meaning that the system will supply power for that duration without external energy inputs (solar or generator). That means, in this case, that we have to be able to supply 3 days of 10 kWh each, or 30 kWh of energy. Deep-cycle lead-acid batteries give the best trade-off between use and longevity when they are discharged no more than 50% of their capacity, so that 30 kWh of energy means we need batteries that store 60 kWh total. For a system this size 48 Volt is the only choice, and at that Voltage 60 kWh works out to 1,200 Ah of battery storage.

We decided to go with Surrette S-1725 batteries. These have a capacity of about 1,600 Ah when discharged over 72 hours (3 days). That is slightly more than we needed, but the other available sizes were not a good fit, and as batteries age their capacity will drop, so these should give the customers a good long life. In this case we went for a single string of 24 of these batteries (each is 2 Volt, and over a 100 lbs!). There are pros and cons for going with a single string vs. two or three parallel strings (one should normally not use more than three strings in parallel), in this case I felt the single string was the best route to go.

We built a battery box out of 3/4" plywood, and insulated it with 2" XPS on all sides (this is an unheated garage). Batteries can handle a great deal of freezing (as long as they are at least 50% charged), but their capacity drops when they get cold. This should make them ride out the temperature extremes in winter (and summer). The batteries actually produce quite a bit of heat when they are charging, enough to keep it warm in the box, even with the garage at -20C. In fact, I'll tell the home owners to remove the top sheet of insulation once spring arrives and leave it off for the summer months. All the charging sources have temperature sensors that track the battery temperature, and they will back off the charging current when the batteries get too warm. Still, it will be better to be able to charge at full current even during the warmer months.

1083

The next picture shows the controller that Magnum uses. While it is not too difficult to work with, there's room for improvement in the user interface. For those that have worked with the other big off-grid inverter brands: It's better than Outback, but not quite as easy as Xantrex.

1082

Note the "88% SOC", that is the State-Of-Charge of the battery, coming from the battery monitor. It gives a very accurate measure of how much energy is left in the batteries, and in this case I have programmed the system to automagically kick in the generator when it goes down to 50% SOC.

There is one more item we installed, because it is so useful for so little money: It's another battery monitor, made by MidNite Solar (http://www.midnitesolar.com/productPhoto.php?product_ID=444&productCatName=Battery%20Accessories&productCat_ID=39&sortOrder=2&act=p), and it uses the battery Voltage to determine the state-of-charge. That is not an accurate method at all, as the Voltage depends very much on the load that's put on the batteries, but it is very easy to read and tells the owners at-a-glance how the batteries are doing. This unit has another feature that makes it worth every penny: If the batteries have not been charged to 100% in the past two weeks a red LED lights up, telling the owners it's time to crank up the generator (unless it's a really sunny day). The number 1 reason why expensive lead-acid batteries die a premature death is because they are left partially charged for too long.

The process of discharging lead-acid batteries turns lead and lead-oxide into lead-sulphate. Now, lead-sulphate is insoluble in water or battery acid. Normally this is not a problem, as the freshly formed crystals are small enough to reverse back during charging into lead and lead-oxide. However, if they are not completely removed every few weeks they will grow, until a size is reached where they no longer revert back, and that particular spot on the plates will now no longer participate in the charging/discharging of the battery. If this is allowed to happen the battery will gradually loose capacity over time, until they are 'dead'. Instead of getting 10 years (as would be expected from these Surrette batteries) it is entirely possible to ruin a good set of batteries in a year by not recharging them fully on a regular basis. Surrette says to get a 100% charge in every 30 days, I believe 2 weeks is a better limit.

1081

This monitor will eventually be mounted inside in a location where it is regularly seen (right now it's mounted next to the inverters, this is temporary). As you can see the MidNite and Magnum monitors are actually in pretty good agreement (both pictures were taken right next to each other). The red LED is illuminated, because at that time I had just installed the unit, so it had yet to see a full battery charge.

I didn't mention the charge controllers we used: There are two MidNite Solar Classic-200 (http://www.midnitesolar.com/productPhoto.php?product_ID=257&productCatName=Charge%20Controllers%20-%20Classics&productCat_ID=21&sortOrder=2&act=p) controllers tucked away right under the inverters. You can see them in the picture next to the MidNite distribution box. At this time IMO these are by far the best charge controllers and the best bang-for-the-buck when it comes to the 'larger' controllers (Morningstar makes great smaller controllers).

Each of these controllers handles 16 panels of 250 Watt each. These are MPPT type charge controllers, so the controller gets around 120 Volt in, and converts that to the 50 Volt (or so) that the batteries need.

As luck would have it (or lack thereof) one of our two controllers arrived DOA. Luckily MidNite's customer support proved to be stellar, and I'm waiting for a replacement to arrive. As soon as that's here we will install it and hopefully get the full array online.

As with the PV wiring, don't look too closely just yet at the wires around the inverters and panel box; we still need to clean that up. The rush was to get the customers into the house, with power, before Christmas (and we met that deadline). We will clean things up in the near future.

That's the whole story, hopefully it is of use. Feel free to ask questions!

-RoB-

Ralph Day

5th January 2014, 06:52

Wow Rob, what a nice project. The only little piece of "learned the hard way" advice I'd give would be to get the customer a good hydrometer and teach/learn together how to use it. Give the batteries an eq charge before too long (for a baseline sg chart), and don't baby those big boys. We started our Surrette's out with insufficient charging (solar,wind and generator), probably paying for that now with some cell replacements and lagging cells in the bank. The high end of Surrette's recommendations would be required for that large a bank...varying the absorb time by season. The head in the battery box is the best way to find out if you're charging enough.

I chart the sg of each cell monthly (1st of the month) and the water consumption (middle of the month). You can see if a cell is lagging behind by sg, or maybe not charging much by the water consumption (that's how I found 2 that needed replacement). Probably not the most popular job off grid, but the most necessary.

As far as 3 days autonomy goes, I think that's asking a lot for the re-charge capability of the pv. I find that 1.5, maybe 2 days is/was our max before supplementary charging was /is required. A few litres of propane for charging is less cost than replacing a pooched battery bank.

Apart from all of the above, I can see the customers have got the best install and customer service possible (Solacity), so they are in good hands.

Ralph

Rob Beckers

5th January 2014, 17:32

Thank you for the kinds words Ralph!

From Surrette's own battery manual I got the picture that these big batteries like to be 'worked'. They recommend deep cycling, especially in the beginning, in their own words "to open up new reaction sites on the plates" as it seems new batteries are still forming and not yet at full capacity out-of-the-box/pallet.

I am giving the home owners a very nice hydrometer as a house-warming gift: I just imported a box of HydroVolt hydrometers from the Swiss manufacturer (http://www.compaselect.ch/battery_tester.html). By all reviews they are the best you can get; made from plastic (doesn't break when dropped), very easy to use and read, very accurate, and automatic temperature compensation. Not expensive either. I still have to put them on our Web site, and figure out a sale price, probably $30 since MidNite is selling them for $40.

http://www.midnitesolar.com/images/products/photos/hydrometer_web.jpg

The charger voltages are currently set to the mid-values recommended by Surrette. Thing is that charging a big bank like this with a generator takes a very long time (the better part of a day!), using up quite a bit of money in propane, so I don't really want to set them any higher than needed. Thank you for the recommendations on record keeping, I will pass that along to the owner and also have them keep an eye on any anomalies between cells (in case we need to raise the Voltages).

Most of the time, most of the year, the solar panels should be able to keep the batteries charged. They have already been instructed in strong terms that the batteries need to see a full, 100%, charge at least every 2 weeks, in case the solar PV doesn't get there. The MidNite battery monitor mentioned in my previous post will help track this, so it is obvious when supplemental charging is in order.

Right now they are running them down in about 3 days, since we have had little sun, and half their PV array isn't working yet due to the faulty charge controller. It also took some time to get the auto generator start to work properly, and the first few charges they run the bank down to the inverter cut-off Voltage. So the batteries are getting a good workout as things are.

The problem with taking a whole house off-grid, with people that are not used to it, and appliances that (although relatively energy frugal) are meant for a grid-tie house, is that it's difficult to predict exactly how it is going to work out. It is better to plan and size the system a little on the large side vs. coming up short and ending up with an unhappy customer.

-RoB-

Ralph Day

5th January 2014, 18:31

Hi Rob
Put me down for one at least. I've an email out to see if someone wants one as well so we can share the shipping. I've read about them before, but didn't want to send to Europe for one.

Once you know the price, pm me and we'll figure out how to pay/ship etc.

Thanks

\ralph

Peter Klaassen

5th January 2014, 20:20

That is a good read for some one looking at a house size system. I agree with Mr. Day about learning to be your own power company. Maintenance is something that can save your batteries and a few dollars. The Surette people indicated to me that for warrantee issues a record of specific gravity and maintenance will go a long way for a prorated cell replacement if necessary. I am tracking my batteries similarly to Mr. Day. I have recently put in a Flexnet DC to monitor the batteries in my Outback system. It has helped my better half understand what is going on with the system when I am away. For those less nerdy about batteries and solar charging, a gas gauge is a big help.
I am wondering if anyone has looked at automatic or semi automatic cell watering systems. I have looked online at some systems of tubes and special cell caps but I wonder how reliable these are. I have a siphon tube and a hand valve for watering the cells for now. I would also be interested in an easy to use hygrometer Rob.
Pete

Ralph Day

6th January 2014, 06:54

Captain Klassen, Mr Day was my father.:)

Today's a good day to test the anchors for wind turbines (also last month).

Ralph

Peter Klaassen

6th January 2014, 12:53

Today we are having wind Ralph. I am missing out however on power production with the tower parts on the ground. They arrived on the 3rd, so the turbine will be going up in the next while. How are you doing with your wind power?
Pete
(Captain on the ship only, Ralph. I am not in charge here at home-as I have been reminded)

Ralph Day

6th January 2014, 14:58

Hi Pete
2013 wind production was 657kwhrs (the most since 2004, our start date). That represents 24% of our total renewable harvest. Generator time is only for exercising, total kwhrs from Hydro One was 1165 kwhrs. That's for charging, loads carried while charging (inverter bypasses loads) and any use in the garage (separated from off-grid power system).

Wind isn't much, but unless we doubled our pv we'd still have a whack of utility or generator without it. The utility time is creeping up as I try to nurse the batteries into old age. Bad treatment at the beginning being paid for now.

Ralph

Rob Beckers

6th January 2014, 19:11

Ralph, Peter, I'll keep you in mind for the HydroVolt hydrometer (I have them, sitting in a box right behind me, but want to use one so I know how they work, and have to sort out a few things before I start selling them).

Ralph, I'd be interested to hear what you think went wrong with your batteries. I'm grateful for lessons learned, so I (or my customers) don't have to repeat them. I've taken note of your earlier comments about working the batteries properly. The Surrette S-1725's are part of their 4000-series batteries (much cheaper than the really big 5000-series, with a 7-year warranty, good value for money IMO). I've sold large numbers of 4000-series batteries to people over the years, most of them S-530's, never heard a bad word about 'm.

The charge controllers (inverters and MidNite Classics) are currently set to 58.8V bulk/absorption, 53.2V float, and 63.2V equalization. Those are the middle numbers from the range that Surrette publishes in their battery care manual.

Where the problem really starts is with the absorption time: Surrette says to use t = 0.42 * C / l. Where t is the time, C the 20hr battery bank capacity, and I the charging current. In this particular case that's t = 0.42 * 1300 / 85 = 6.4 hours!!. I have this currently set at 4.5 hours or less than 25A going into the batteries. At the same time the Magnum inverter is also set to charge to 100% SOC when on the generator. I've yet to figure out which of these settings takes precedence.

Peter, in truth I'd be scared to install an off-grid system, especially something as large and expensive as this, without a real battery monitor that tracks State-Of-Charge (like the Outback Flexnet). Unless the owners are really into the nitty-gritty of batteries it is hard to get them to accurately track the state of the batteries, and make sure they get recharged properly. The SOC monitor takes all the guesswork out of it; they drain it down to 50%, the generator kicks in, and it gets recharged beautifully to 100% SOC without any human intervention (other than keeping the propane tanks filled). I'll have to explain to them how to do an equalization every 2 or 3 months, and to monitor and water the batteries, but at least I don't have to worry that the batteries get drained too far too often or not recharged properly.

I've installed the MidNite battery monitor in addition to the SOC system so the home owners can see at-a-glance where they are with the batteries, and plan accordingly what loads they would like to use that day (ie. sunny day, some juice left: Run the laundry!). It will also tell them when it's time to kick in the generator manually to get a full charge into the batteries (the PV alone tends to short-cycle the batteries without ever getting back to a full 100% SOC). The MidNite meter is very cheap compared to all the other hardware, really a no-brainer in deciding to install it.

-RoB-

Andy Rhody

6th January 2014, 22:43

Great work Rob. Those people couldn't be working with a better guy!

Although you said that this residence doesn't have much wind, as you know I have 6.8 Kw of solar and 2 Kw of wind with only a 23 foot tower but still with the storm today, my smaller windmill beat out my solar. 8 Kwh of solar and 11 Kwh from wind. Although we all know this, it's still amasing to see how well wind picks up when the solar slows down.

Wind needs to get cheaper!

Steven Fahey

6th January 2014, 23:04

Ralph Day

7th January 2014, 12:02

Hi Rob
I'd be happy to use a HydroVolt and compare it's performance to the good glass hydrometer I have. There's years of records (and experience using it), so a good comparison should be available. Sell me one, please!

We started out in a deficit charge mode...the generator and MX60 absorb voltages and times were the minimums recommended, which was not enough. 58v, 2hours, and without a hydrometer to know state of charge we ended up with some degree of permanent sulphation. Don't do that. Start off with a full charge and eq until the sg's don't rise any longer...get sg's for all the cells to act as a benchmark for each cell.

I used Word to make a chart with the date and sg for 24 cells (also one for water addition). The end of the month for sg and middle of the month for water..that way you're not taking sg readings when the acid is most concentrated or diluted by water addition. It was really advantageous to have that information when I dealt with Surrette on the 2 cells replaced under warranty.

I built a 2amp charger which I can put on one cell at a time (only the lagging cells). Eq charging is every 3 months. In the summer I drop the absorb voltage and increase the time, winter i increase the voltage and never reach 3 hours.

The charge numbers are more of a guidline. Each cell has different characteristics, so the best/only way to tell if you're fully charged is with an sg reading. A run of sunny days and you might not need to absorb as long, I know I don't . \that's where regularly reading your pilot cell comes in. If it's been cloudy for a couple of days one full absorb might not fully charge the batteries. If it's been sunny for a couple of days might make a long absorb just a water and energy waster. Time and training will settle it all out.

I wouldn't rely on a batery monitor for anything but a cursory look at soc. If it looks like you're near your needed generator point...check the sg and see if you really are (you might be lower). I used to let things go to 25 or 35% depth of discharge, but an sg check showed 50%. I don't trust meters.

Good luck to you and your lucky customers.

I need 2 HydroVolts when you are ready to ship.

\ralph

Rob Beckers

8th January 2014, 08:27

If I may butt in...

Have you tried a Generac like that one on natural gas? The spec sheet doesn't say anything about it, but the website only makes a vague claim.

Very nice install. If I had my druthers it would look something like that (but with sun tracking) :)

Steven, the official word is that the 6kW EcoGen is the only Generac that they do not have a conversion kit for. It is only available for LPG. That said, it may well be that someone makes the proper jets for NG, or a conversion kit, so it can be configured for natural gas. Not supported by Generac though.

-RoB-

Rob Beckers

8th January 2014, 13:26

Ralph, since you are familiar with the Surrette batteries: What are your set points (Voltages) for them? I've set low-Voltage cut-off for the inverters at 44.0 Volt (1.83V per cell); according to Surrette's own documentation that should be around 20% SOC. However, the home owner tells me that when the system cut out yesterday (he had switched off the generator, so it did not automatically kick in), it came back up after the generator was started at 68% SOC on the battery monitor. I don't have an explanation for that.

Voltage wouldn't be very precise as a measure of SOC, but it can't be this far off. The Magnum battery monitor automatically calculates battery efficiency, I am not sure what their algorithm is, all I have to give it is the bank capacity (about 1,600 Ah for the 3-day discharge rate we expect here). It has had a number of cycles to refine its settings, so SOC calculations shouldn't be off too much any more at this time.

-RoB-

Ralph Day

8th January 2014, 14:45

Hi Rob
My low voltage cut off is also at 44v, I think it's everyone's default (inverter manufacturers). Cut in is 52. Mine has never cut out.

Your customers might have been in a situations where the soc was low and with a big load like a submersible pump or furnace fan the "voltage sag" put the voltage below the LBCO setpoint even though the state of charge was above where that theoretically should happen. If the generator was set to operate automatically (with those 5 15 and whatever minutes setpoints) the system would probably have started the generator and the shutdown would not have occured.

The return to activity by the inverter after the generator was running was due to the LBCI Low battery cut in setpoint. Once that is reached the inverter resumes function. Perhaps the inverter operates in charger only mode after an LBCO event..to ensure that charging to a minimal level is achieved before the homeowner is allowed to use any power. That level would also be reached by the pv input charging, but if it's only 1/2 operational the generator did it faster (or it was dark?). There's nothing like the feeling when the lights go out, and you are your own power company!:sad2::cry:

You can check the LBCO and LBCI setpoints and change them to better suit the clients, or better yet, they can. There's little or nothing you can fry with the programming and setpoint changes (short of letting values go too low or too high). If you make a boo boo you just set another point or go back to manufacturers defaults.

I really think it was a function of low soc, voltage sag perhaps, and the generator not being on automatic function. Mine is never on auto function...it's diesel and too many things can go wrong (plus I have Hydro One as my multi billion $$ generator). If the soc was low enough to shut down the inverter it must have been quite low. Remember the full charge and eq for commissioning. And don't rely on expensive battery monitors exclusively...the $30 HydroVolt is a much better thing to rely on for state of charge monitoring. ;)

Ralph

Rob Beckers

8th January 2014, 17:54

Thanks for the info Ralph!
Actually the Magnum default LBCO is just 40 Volt. According to the Surrette charts that would mean a completely dead battery!

The 44V choice is so they have a little extra wiggle room in case the generator isn't working (for example) before the lights go out, without doing actual damage to the batteries (they'll quite happily do 80% discharge, but doing it all the time will give you less cycles over their life). Normally the generator is set to automatically kick in at 50% SOC, but if they switch the generator out of 'auto' of course all bets are off..

By the way, for the Magnum inverters the battery Voltage has to go below LBCO for a full 60 seconds before the inverters kick out. That should avoid nuisance drop-outs due to heavy loads coming on. The Magnums do not seem to have a particular LBCI set point, not sure what the policy is on that.

I'll go over there in the weekend and do some SG measurements of the cells to get an idea of what's happening between SOC and actual.

-RoB-

Rob Beckers

11th January 2014, 16:16

I'm continuing to struggle a bit with the battery bank, and trying to make sense of it all...

Thing is that while we sell quite a few batteries and off-grid supplies, it's fairly rare for us as a business to do an installation. Those we do are usually small, ie. for a cottage, just an inverter, charge controller and battery bank with minimal accessories. A large installation like this is fairly unique, with all the bells and whistles. Normally we're either not involved in setting things up (programming inverters etc.) beyond some general advise, and of course, I don't live off-grid myself so seeing a full-size household like this off-grid and guiding the home owners through the tribulations that come with it is a first for me.

That said, I am very much willing to learn! (and in case anyone thinks this is at the expense of the customer: They are getting a very good price, and very likely much better service than anyone else would have given. I've spent many hours over there!).

I was over at the site again this morning, the plan was to measure specific gravity, take some Voltage readings, and try to figure out why the generator was auto-starting (due to the Voltage threshold set in gen-start module of the inverter) while SOC was still showing around 70%.

To start with the SG readings: I tried the new HydroVolt hydrometers we just received in from Switzerland. They are supposed to be amongst the easiest and most precise out there. Here's what I got:

1087

In case that's not too clear, here is a closeup:

1088

The SG is literally off the chart (on the low end)! In fact, this would indicate a dead battery, and since they were humming along nicely that just couldn't be.

To make sure I tried another hydrometer (I have a few dozen of them in stock), same result. Just in case the acid was stratified I filled/emptied the meter a few times, stirring it up that way. There's no magic to them, the big black part is a bladder, squeeze it to get the air out, stick in the acid, and let go, it'll fill up the chamber with the two freely rotating disks with acid. One of those disks is in essence a float, to indicate SG. The expansion coefficients of the two plastic disks are such that they cancel out any effects of temperature, leading to automatic temperature compensation. Reading it out is by simply holding it up vertically.

At this point I don't know what's happening. Did I get a bad batch of hydrometers (this company makes hydrometers for many different fluids, not just battery acid, maybe the wrong type of float was installed)? I've sent a message to the designer, let's wait and see.

I measured the Voltage of each cell with a (rather expensive, and hopefully accurate) Fluke multimeter: They were all very close to 1.980 Volt (the range was 1.956 through 1.996 Volt, but the lowest cell would measure a 1.975 a moment later, it's very sensitive to the load at that moment). According to Surrette's own literature that places SOC somewhere around the 40% - 60% mark:

1089

The SG at that point should have been around 1.175. By the way, the load on the batteries was moderate at this time, around 15A, though this various (rapidly) between a few Amps and 40+ Amps when the furnace kicks in.

That is one part of the mysteries-to-be-solved. The other is that at this Voltage the battery monitor was indicating 80% SOC. The same monitor showed a deficit since the last full charge of 300-some Ah, which makes sense in that the 3-day capacity of a set of S1725's is supposed to be 1,600 Ah (1,300 / 1,600 = 81%).

Now, if the Voltage indicates 50% SOC that means with a (say) 350 Ah deficit since last charge that bank is only good for 700 Ah!! :eek:

Surrette says that it takes some time for a new bank to attain its full capacity, it needs to cycle through a number of discharge/charge cycles to get the full surface of the plates to participate in the chemical reaction. This bank has probably 10 or so cycles on it (some deeper than others, most around the 50% SOC mark I'd guestimate). If so, when can one expect to see full capacity? The bank has not been abused, with the way LBCO is set it should stay above 20% SOC at all times.

That also leads to the question of how to get that battery monitor to indicate accurately. Magnum claims it has an algorithm build in that automatically determines charging efficiency. The bank capacity is a setup parameter though, and likely everything is based on it being accurate.

I understand that I'll have to first resolve the SG measurement issues, so I can take accurate readings. That being the only good way to determine actual SOC, so I'll focus on that first.

We're also replacing one of the two MidNite Classic-200 controllers; one isn't working right (didn't work from the start). The replacement just came in and will be installed Wednesday. Once we have both up and running I can do an equalize on a sunny day without breaking the bank in propane use (the two controllers and the 8kW array can deliver nearly twice as much current into the battery bank than the generator can).

-RoB-

Ralph Day

11th January 2014, 16:42

Hi Rob
Sorry to hear about the battery woes. If you're worried that the HydroVolts are not working properly, go to a Carquest, Canadian Tire or Napa store and try to find a glass float type hydrometer. I've found that even the cheapest (glass float in tube, not plastic pointer type) ones will read the same as my good one if the float doesn't stick to the side of the tube.

The electrolyte might be stratified so that the topmost measureable liquid is the weakest. Has an eq charge been done yet? Has there be good active bubbling during the absorb phase? You need that to happen to cause some stirring of the electrolyte. Gassing is necessary, even though it is really a sign of wasted energy. Without it, especially with the tall single cells, you can have stratification occur.

I didn't have an hydrometer when I started out, so I can't say whether my batteries exhibited the same behaviour or not.

Try a HydroVolt on a forklift, car or truck battery at home or the present office/wharehouse. If it reads poorly on a known state of charge battery, then maybe it's the units themselves. Do you have a regular auto garage you go to? Could you take a HydroVolt unit there and check a battery of know soc?

1. Determine whether HV units read correctly so you can measure the bank's sg's
2. Do some heavy generator charging on the bank...bubbling. Absorb then eq. Once sg's stop rising you have a baseline sg value for each cell.
3. Cross fingers

Best of luck. Steep learning curve, non?

Ralph

Rob Beckers

13th January 2014, 09:13

Oddly all the automotive batteries I have are 'sealed' (they happen to all be NAPA batteries); they are regular lead-acid, but the tops are a single piece of plastic that apparently clicks in place after they fill them at the factory, so it cannot be removed. I tried. They are advertised as "maintenance free". So no luck using those to measure SG.

I came across a post on the Internet that stated how equal parts by weight of water and regular sugar will yield a solution with an SG of 1.2317. My Ikea kitchen scale is accurate (or at least repeatable) within 1 gram, so I measured off 100 gram of sugar, and 100 gram of very hot water. The sugar easily dissolved in that little water, and lo-and-behold, the HydroVolt measured exactly 1.231! That third digit after the decimal point is up to interpretation, as this approaches the limits of resolution on the unit. It doesn't matter, all I wanted to find out is if it's working, and clearly it does!

I rinsed out the hydrometer very, very, very thoroughly with hot water, shaking it, to make sure every last trace of sugar will dissolve and flush out, repeated that a number of times, and gave it a final couple rinses with distilled water. It should be good to go again for use on batteries.

So, that at least brings some clarity. The HydroVolt is fine. The SG measured from those cells really is that low.

What I think is happening is that the acid is badly stratified. These are very tall cells! The straw on the HydroVolt is just long enough to reach the top of the acid. I will measure again after bubbling up the acid thoroughly during a charge cycle.

Since it's clear the acid is stratified, I will raise the absorb Voltage to 59.5 Volt for now (which is still below the upper limit that Rolls prescribes). From what I've been reading these cells don't need that much during normal use, but when new they have a hard time 'breaking in' and getting the SG up. As soon as I have that second charge controller working I'll also do an equalization charge (on a sunny day, starting with the generator in the morning to get a bulk charge in).

FYI, the cells do bubble audibly and visibly while charging with the absorb Voltage set to 58.8 Volt, the middle of the range prescribed by Rolls. From all I've been reading and seeing I just think they need to be worked a little harder in the beginning of their life.

In all likelihood part of those plates in the battery cells is not yet participating in the charge/discharge cycles, which would explain the sub-par capacity the bank seems to have at this time. This jives with other stories on the Internet about these batteries: They seem to need a good kick in the rear (or a couple of kicks) and a number of cycles before they behave as they are intended (one person mentioned 50 cycles!).

These are still new batteries, and they have not been sitting very long in this state, so I'm pretty confident no damage has been done. It's a lesson though, these really need a good charge, and better yet, an equalization, to start off. Probably not a bad idea to do an EQ every 3 or 4 weeks for the first couple of months.

Yeah, there is a learning curve for sure, and maybe these large/tall batteries have a little steeper curve than the usual L-16's. I now also better understand the saying that most people who are new off-grid life will ruing a perfectly good set of batteries within a year. In this particular case the home owners are not quite ready yet to take care of the batteries themselves, and I'll try hard to make sure they get taken care off properly until they are. This is a lot of money in batteries, and I don't want anything to happen to them! Not so easy since I don't live there...

-RoB-

Ralph Day

13th January 2014, 15:35

Hi Rob
I think I recall hearing about the HydroVolt needing an extension to reach the acid on some re batteries. Factory or homemade I don't remember. If the well is kept 1/2 to 1/4 inch from the bottom of the well that's the sweet spot.

If you check with Chris Olson you could find out a lot about Surrette charging parameters. He has his new bank really cooking, but sounding healthy. The customer's will take a lot of charge to get going...just hope the shortage of propane doesn't put a crimp on what's needed.

I once went to Canadian Tire and asked to have my hydrometer checked against theirs and a sample of the bulk electrolyte they had on hand. Small town courtesy, they said sure. BAng on 1.265 from their sample. But you know your HV's are fine, that's good.

Is the full array in operation yet?. I decided to keep the batteries here on float (Hydro One power) and do my 3 month 2hour maintenance eq charge. I have to do it when Janet's not here...she can smell when I've eq charged even if I don't open the battery box.

Ralph

Peter Klaassen

14th January 2014, 06:55

Hi Rob,
With the system you are dealing with do you have the ability to monitor power in and power consumed? Looking at these numbers might give an idea what is up with the batteries. The flexnet dc I have here is somewhat hard to fully understand with the de-rating used for generating charging. As with your system, the input figures from the battery bank have to be accurate. You are using the 20 hr rate for your numbers? For us here there has been little sun so the generator is getting significant use. Over time the battery monitor is supposed to measure the charge efficiency but it is just saying 100% which I don't believe. The batteries are coming on 8 years old. I would agree about the possible stratification. One of the characteristics of solar charging was that it was slower and easier on batteries. This may not be true with modern charge controllers and an 8 kW array. In our neighbourhood however we have had little sun lately and on cloudy day I get next to no production.

Rob Beckers

14th January 2014, 07:34

Peter, The Magnum battery monitor shows Ah in and out, including since the last full recharge (or what it considers a full charge). It calculates battery charging efficiency automatically, they don't explain what the algorithm is. The one parameter that needs to be set is the size of the battery bank, in Ah. I had this initially set to the 3-day (72-hour) value, since the bank is sized such that without charging sources the home owners should last about 3 days. It's clear that at this time the bank does not reach that capacity (yet). SOC by the monitor is simply calculated (near as I can tell) by subtracting the deficit from the bank capacity and dividing that by the same capacity (ie. 1,600 Ah battery bank, -400 Ah deficit since last recharge, SOC = (1600 - 400) / 1600 = 75%).

The 8kW PV array can generate 140A into the batteries. That's more than the generator can produce. Even on a cloudy day I often see half the rated output, so PV is important as a charging source here.

Ralph, the full array is not yet working; we're hoping to get the (second) replacement for the MidNite Classic-200 today. If it arrives we'll install it tomorrow. Until then only half the array is working.

-RoB-

Peter Klaassen

15th January 2014, 10:16

So Rob, I am wondering about entering the battery bank capacity on the Magna Sine monitor at the 72 hr rate. This is something that seems to be non-standard for the amp-hour measurement and might be causing errors. I looked at the Monitor manual online and it would seem to me that the amp hours for the battery bank should be entered at the 20 hr rate. For the Flexnet this is the case and the system is reset (turned off and back on) when one knows that the batteries are at 100% charge. The battery monitor would seem to have to work with standard values and I think the 20hr rate is the one most commonly used.
I am not sure I am helping - I would guess that you have been in touch with the magnasine people.
Pete

Rob Beckers

15th January 2014, 18:36

Peter, yes, I did ask Magnum support. Unfortunately the support person I am dealing with has a habit of answering everything but the questions I ask. I've somewhat given up (I ask "how does it calculate ...." and the answer I get is along the lines of "most people seem to do this").

Anyway, there's no magic in that battery monitor: It bases SOC on what you set as the battery bank capacity, minus the Ah's that have been removed since the last full charge (and plus what's been put in from charging sources). Since this system is designed to run 72 hours without any charging sources on the batteries, with the loads we have, it would follow that the battery capacity is what Rolls says those batteries can deliver in Ah over a 72-hour discharge period. In fact, we only discharge them to 50% over 72 hours, so the discharge rate is actually less than the Rolls published rate.

I was over at the site today to wrap up some installation details, put wiring in conduit, and make it all look pretty. Also installed the remote control for the inverter, and the MidNite battery monitor in a place where the home owners can see it without going into the garage (where the inverters are). The batteries were at just about 50% according to Voltage, so I fired up the generator and told them it really needed a full charge this time. I've raised the absorb Voltage, hopefully it's stirring up the electrolyte a bit more now. Got the second charge controller replaced and working as well; when the sun came out for a moment late afternoon (overcast here all day) it immediately was doing 90A into the batteries between those two controllers!

I'll arrange for doing an equalize next. Maybe over the weekend.

-RoB-

Bryan Couch

15th January 2014, 20:02

Ralph Day

16th January 2014, 08:08

Rob Beckers

16th January 2014, 08:33

I've been following this thread since you started it, it's just the kind of thing I'm interested in! Great write up.

I've been waiting while you sorted out the battery issues to comment and since it seems you're close I have some questions you may or may not have answers to.

1 - This house was not designed to be off grid... What kind of changes would be involved if it was originally designed to be off grid?

2 - What size is the house (sqft, bedrooms, bathrooms...) and what size is the family (0 kids, 2 kids)?

3 - Is it a typical built to code house or have they gone beyond the minimum for a more efficient home?

4 - What kind of changes did the family have to make to get their usage down to 300 kwh/day?

Anyone with ANY input on these questions feel free to chime in, the more info and ideas the better. :)
We still haven't got a yay or nay from the mining company yet but I'm so interested in the passive house, off grid "eco" construction possibilities I can't seem to stop researching and reading. The ideas and technology out there but not being used is amazing!

Hi Bryan,

1) If it was up to me, and I had to design a house to be off-grid from the get-go, I would have opted for a design that utilizes passive solar for some of its heating, and use a wood stove as the main heat source. Maybe a regular (propane) furnace would be part of the mix but I would try to do without having to use it for most days. Even though the furnace runs on propane, the blower motor is a very large load!

Originally this house was going to have air conditioning as well. That had to go; it's just too large a load to run comfortably off-grid (or you'd have to size the system even larger).

We had to clear out a bit more land in the back to place the PV array (that wasn't planned). It would be good to plan these things earlier on, not just PV modules, but where to place the battery bank (it's big!), inverters etc. It worked out in this case, but that was in part just luck.

In truth though there's not much of a difference (I believe) between a grid-tie or an off-grid house as far as the structure is concerned.

2) Not sure what the overall size is, probably around 2,500 sq. ft. I believe it's 3 bedrooms, 2 bathrooms (I've mostly been working in the garage where we installed the electrical power plant). The family is 2 adults, 2 teenagers (the latter are bad when it comes to electricity use! ;) ).

3) Not sure what residential code is these days. The house is very tight and very well insulated. That's a good thing, especially off-grid (so that blower motor doesn't have to run too often).

4) We were just in time to look at the furnace, HRV, fridge, dryer etc., and chose energy efficient models. Meaning DC motors for the furnace and HRV, propane dryer (though the motor in there still draws quite a bit!). They are trying hard to keep use down by diligently switching off the lights when not needed, and I believe they'll also get in the habit of scheduling tasks according to energy availability (ie. do the laundry when a sunny day is forecast).

I believe that others that live off-grid 'grow' into this more gradually, starting with a small system with very minimal demands, and adding onto this as needed. It's fairly rare to have a household that's thrown into the deep end, going from grid-tie to off-grid in a day. Keep in mind though that 300 kWh/month is a pretty generous budget when it comes to off-grid.

-RoB-

Rob Beckers

16th January 2014, 08:44

Hi Rob
Are the owners shutting down the generator when they've heard enough, or are they worried about running out of propane? They really need to realize that their big investment in those batteries might be lost or compromised if they don't get them charging properly.

Don't forget, they should be at a full soc before an eq charge is done. I do a 2 hour eq charge every 3 months for my set. It's better than having to do the recovery eq charging. Perhaps you could get them to run the generator in the morning before you arrive so the bulk of any charging is done, you just get to fine tune the system.

Once the sun comes back they should be able to handle charging with the pv available. But if the soc is low, and you have a sunny day coming up, a start with a couple of hours in the morning with the generator will ensure that a full soc is reached with the pv during the rest of the day.

Ralph

At this point the home owners shut the generator down when they see 90% SOC. I've told them that's OK, as long as the batteries get a full charge once every 2 weeks minimum. The concern is the propane cost; prices have jump 50% (!!) in the last couple of months. There's a propane shortage going on.

The issue with shutting down the generator prematurely is that the gen-controller sees it, and it sets an error code. With the Magnum inverter that means it will not switch back to auto-generator start, until that fault has been cleared. I've shown the home owners how to do this, but that has not sunk in yet. Trouble is that without the auto-gen start the system will run down until LBCO, and the power will go out suddenly.

I still have to sit down with the home owners and really go over how to treat batteries and what's involved. So far I've been telling bits and pieces, in the hopes that conveying the facts gradually will help with the steep learning curve.

It really is not plug-and-play to live off-grid with batteries (maybe it would be more so if AGM or gel batteries were used, but those come at 2x the price and 1/2x the life expectancy of regular flooded lead-acid). That is a message I will convey even more strongly to prospects in the future. It may cost me a few customers, that will go with the competition because they promise an effortless off-grid life, but I'd rather have customers that know what they are getting into.

A good EQ is the next thing I'm going to set up. Let's see if we can get that SG up to normal values. I've also purchased some vinyl tubing so I can put a longer straw on that hydrometer, and suck acid in from slightly deeper, that could be skewing values as well.

-RoB-

Ralph Day

17th January 2014, 10:28

Hi Rob
Goodness knows, you're going above and beyond the retail call in this project, but, if I might suggest, how about pointing the owners in the direction of the NAWS forum? There's a wealth of off-grid information and experience there that they could benefit from. Not to disparage what you're doing for them, but to avoid re-inventing the wheel they could find out a lot to help them along on that forum too. There's just that much more experience with a larger pool of people to benefit from. You and I can only tell them/show them so much.

Ralph

Ralph Day

22nd January 2014, 07:36

Hi Rob
I scarfed this quote from the NAWS forum. Link as well, if you'll clear it.

"A battery monitor is often compared to the gas gauge in your car, but that is not a good analogy.

The gas gauge in your car actually measures the level of gas in your tank. A battery monitor doesn't measure the level of anything. It is more like the odometer in your car. The odometer counts miles and the battery monitor counts ampHours.

Suppose you have a car that gets about 25 mpg and has a 15 gallon gas tank. Suppose also that your gas gauge does not work. What do you do? You use your odometer. For example: After a fill up you drive 150 miles and you expect that you have used 6 gallons and have 9 gallons remaining in your tank.

But what if you do not completely fill your tank. For example, starting from a full tank you drive 150 miles and then you add 3 gallons to your tank and then drive 100 miles and then add 4 gallons to your tank and drive 150 miles and then add 5 gallons to your tank and drive 100 miles. At this point you estimate that you have 7 gallons remaining in your tank, but that estimate is not too accurate because your mileage is not ever exactly 25 mpg. The only time you know exactly how much gas is in your tank is when you have just filled it up (or when you run out of gas).

So it is with your battery monitor. It can be very accurate counting the amphours in and out of the battery, but it can only estimate the state of charge based on your assumptions of the battery capacity and of the battery efficiency. It can read 100% (by counting amps into the battery) before the battery is fully charged. The only time it is exactly accurate is when you fully charge your batteries and reset the battery monitor to read 100% full.

The trimetric can automatically reset to 100% based on criteria that you set. For example, you may have decided that your batteries are charged when the voltage is at least 14.4 volts and the current is below 20 amps for a minute. But is that really when your batteries are charged?

If you're going to train your monitor how to recognize a fully charged battery, then you yourself must know when the battery is fully charged. You may accomplish this by monitoring end amps during absorb charging and using an hydrometer or refractometer to know when your batteries are fully charged.

Bottom line: the longer its been since a complete charge, the less you can trust your battery monitor. And you can't trust it at all if its not set up correctly. That said, battery monitors are valuable tools in estimating your SOC, but you have to understand their limitations. You must have a hydrometer or refractometer, and and a way to monitor the current into the battery.

When you know the 'end amps' current and absorption voltage that represents fully charged, you can program the trimetric to be more accurate."[/I]

http://www.wind-sun.com/ForumVB/showthread.php?22359-batteries-dead-or-not/page2

Ralph

Rob Beckers

22nd February 2014, 17:34

It's been a some time since my last post, and since I was last at the customer's site. For a number of weeks I have been trying to make an appointment with the homeowner to do an equalization, and to teach him a bit more about batteries. That finally happened today. The last week the guys were out there finishing up the PV array wiring; the new rules require the wiring to be in conduit when less than XX feet above the ground. We put it in liquidtight (the flexible gray stuff). Final inspection by the ESA inspector will be scheduled for next week.

Before continuing my saga about the battery bank, here are a few pictures of the installation with everything said and done. These were all taken today. First one is of the array, turning sunshine info glorious Amps!

As you can see the array is sitting high enough off the ground that the snow doesn't cover the bottom (quite a feat with all the snowfall we have had here!).

1092

This one shows the backside of the array. You can see the liquidtight conduit with the wiring. This shows better how the seasonably adjustable racking works too. They go from 30 to 45 to 60 degrees. Look very closely and you can see the two MNPV6 combiners halfway down the row.

1093

Now that battery bank that's giving me so much grief. It also shows the inverters and charge controllers. In all I think we did a pretty neat job! By the way, note that big conduit on the left side of the far end of the battery box; that's the hydrogen vent, goes to the outside.

1094

A close-up of the electronics. All humming along nicely. A total of 8.8kW in inverters. It is working closely as I calculated: Most of the time it is running on just one inverter, and we offered that option to save some money. They would have to watch their loads very closely though to stay within that 4.4kW budget. Instead they opted for the dual-inverter, and that second inverter does come on regularly. It essentially means they can run anything they want anytime they want. Makes life much easier. Of course, there is still the solar budget to consider, so they don't have to run the generator too much.

1095

The two MidNite Solar Classic-200 controllers take care of the solar array. In total they deliver up to (around) 140 Amp into the batteries!

The small control panel of the right hand side is Magnum's ME-RTR. It's needed when inverters are stacked, and the 'brains' of the system. It has to be mounted within a few feet of the inverters, which is not so convenient for the home owners.

1096

For that reason we installed a regular remote controller, and a MidNite battery monitor inside the house. So they can see at-a-glance what's happening with the system, and in particular with the batteries.

1097

They can also control certain things, such as the generator, from this remote controller.

As you can see from the ME-RTR picture it was doing equalization, at about 63.5 Volt, with 108 Amp coming from the generator into the batteries (today was partially cloudy, the charge controllers were helping but they couldn't keep the Voltage up high enough for equalization by themselves).

We are essentially finished with this project. It has been turned over to the homeowners, and for us the only thing left is final electrical inspection.

That is it for this post. I will post the current battery situation in the next post, after some more coffee!

-RoB-

Rob Beckers

22nd February 2014, 18:46

So, back to the batteries...

I should preface this a bit: The cost of propane is now double (100% rise) over what it was at the start of this winter. The homeowners have therefore been somewhat reluctant (understandably, after they got the last refill bill) to start the generator, and set the system on "manual", so auto-start wouldn't kick in. In part the confusion was that the battery monitor would still indicate 70% SOC-or-so while the gen-start would try and start the (propane) generator. The gen-start works off of Voltage though, and it was set at 45.4 Volt with a 30 minute delay (Voltage had to go below this for 30 minutes before the genset kicks in, to avoid starts due to transient loads).

I have since learned that SOC battery monitors are really very simple things that are only as accurate as the Amp-hour rating for the battery bank that one puts in. It is clear that the battery capacity of this bank is falling short by some-and-a-bit, so the SOC indication is overly optimistic at this time (I have it set to 1300 Ah currently).

Part of the explanation seems to be that these cells come from the factory only partially formed, and need a number of cycles, and some very thorough charging, to get them up to snuff. That was also the finding of a thread on another forum (http://forum.solar-electric.com/showthread.php?21330-Commissioning-Day&highlight=s-1725) where someone was trying to "break in" these Surrette S-1725's. The need for a number of (frequent) equalization cycles, (initial) high absorb Voltage settings were cited.

The batteries have not been equalized since commissioning two months ago, today was the first time. That should have been done earlier, but despite attempts life worked out this way.

In short (from the above), I think this bank has been getting discharged a bit more than I would like, and it is not clear how well it got charged after these events. In retrospect, the very low specific gravity readings were likely a combination of actual low SG, and in part some stratifying that was going on because they did not see a good (bubbling) charge on a regular basis.

These days the sun is much higher on the horizon, with more hours a day, and thanks to the large (8kW) array the batteries should get charged fully on a much more regular basis. Without the need for running the genset. In fact, it should now be pretty rare for them to go a week without seeing a full charge just from the sun.

I had already set the charge controllers some time ago to take the Voltage up to 59.5 Volt for absorption, a bit higher than is usual, in hopes that this would reverse any sulphating that could be going on.

Prior to equalizing the generator ran all morning (about 5 hours) to get a full, regular, charge into the batteries. After about 2.5 hours of equalizing today at 63.5 Volt the SG of the cells still read on the low side: Some were around 1.22, most around 1.20, but a few cells were still around 1.70 (a full cell, as new as these are, should read around 1.26). The cells were all bubbling vigorously (we took the caps off). Distilled water was added where needed as well (they have not been using much water so far).

At the start of the EQ the SGs were reading in the 1.16 range for most cells (much too low for cells after what should have been a full charge), some lower, so they are clearly responding. Not quite as well as I would like though.

I understand that only so much can be done in a single EQ cycle; it takes multiple regular discharge-charge cycles with a few more EQs thrown in to continue to raise the SG of a full cell.

The conclusion at this point is that we have some sulphating going on. Given the age of the cells, and since we haven't really abused these cells too bad, that should be reversible. I have advised the homeowner to repeat the EQ at a once-a-week schedule as the opportunity arises (two sunny days in a row will do it, one to get the batteries up to a regular charge, so it doesn't take long the next day to top them off, and then start an EQ around noon the second day).

Another conclusion I want to emphasize is that it is important to get a full charge into the batteries before putting them to first use. We failed to do that (being in a hurry to get everything hooked up so the homeowner could get off the generator that was running 24x7, and with -20C in that garage while running out of time on a late afternoon).

The homeowner is now also (finally) reading the Surrette battery manuals, and I have directed him to the Arizona Wind & Sun forums. That is giving him a much better understanding of how batteries work, and he understands the need to get the SG up to reasonable values. So, I am hopeful that between the fairly high absorb Voltage and his EQs we will yet get these batteries in shape. He also understands now that if the auto-gen-start says it's time to run the generator it really is needed (with the amount of sun we have this time of year that shouldn't be too often anyway).

If anyone has suggestions on how to proceed with these batteries that will be much appreciated!

-RoB-

Ralph Day

23rd February 2014, 07:04

Hi Rob
The saga continues...
Just a reminder, eq charges should be done on a full SOC battery. Sg of 1.265, then eq for a couple of hours (for a maintenace eq charge). For a corrective eq charge you have to monitor the electrolyte temperature and sg readings closely as time progresses (all in the Surrette bulletins). But 2 hours of eq when the starting sg is below full SOC is really just a higher voltage absorb charge, with extra bubbling.

Scary about the cost of propane. I'm eeking out the last of my tank from 2012 by using a one burner electric hotplate for kettle heating when there's enough sun (or Hydro One is running the house loads). Just hoping I don't have to call for a fill before the heating season is over and the price drops to "normal" levels. Who knows what normal will be after this winter.

Sounds like the homeowners have gotten the best setup and send off from Solacity that anyone could ask for. The rest is up to them. Wouldn't it be nice to have an energy storage medium that wasn't so twitchy and susceptible to degradation as FLA batteries? The news and forums are full of vapour-ware products, but the centuries old lead acid battery still carries the bulk of energy storage for us all.

Ralph

Rob Beckers

23rd February 2014, 08:11

Not sure why propane would have to be so expensive, or why prices are as volatile as they have been this winter: The stuff is skimmed off natural gas, and North America has natural gas coming out of its ears at this time...

I realize that an EQ should come on top of a "full charge" with SGs at 2.65, but I can't get them there no matter how hard we try at 'normal' Voltages. The charge controllers are already set at 59.5V bulk/absorb, and I've switched them from fixed-time absorb to 1% of C-20 (13A) to indicate end-of-absorb. That's about as high as I dare go for regular charging, in hopes of not doing damage. The EQ Voltage/charging is needed to get the SGs to rise to more regular values, though right now they seem to top out at around 60 - 75% SOC when reading the SG values from various cells.

All the chargers have temperature sensors and compensate for cell temperature (and I assume the software is smart enough so they back off when temperatures get unhealthy). I also opened up the battery box while doing an EQ. With around zero ambient temperature they did not get all that warm (holding a hand alongside indicates about luke-warm or 20C for the cells during the process).

On another note: Ralph, how do you like the HydroVolt hydrometer? How does it compare with your other hydrometers in accuracy and ease of use? I have not had or used other types, but after reading all the complaints on how hard it is to get a good reading from a float-type hydrometer, or how long it takes to read all the cells with a refractormeter type meter, it seems a breeze with this meter. It's fast, it's repeatable. The only thing I found is that after rinsing with distilled water it takes a rinse with the acid of the first cell before getting a good reading.

-RoB-

Ralph Day

23rd February 2014, 10:03

Hi Rob
The Hydrovolt works well. The readings are the same as the good glass hydrometer I have. It's just getting used to reading a rotating scale instead of a vertical scale. I did find that the first reading needs a couple of squeezes of the bulb to get a real reading.

My son in law explained the propane problem as explained to him: It's all in the trucking.

All ontario's propane comes out of Sarnia with one loading depot. A truck can fill only so fast. There are up to 18 trucks waiting to fill at any one time. A commercial driver is limited to how many hours they can be "driving"/on duty, in a day/week/month. I think the time spent waiting in Sarnia counts as on duty. Drivers will have to stop along the route to far eastern Ontario when their 14 hours for the day runs out. Slows up the delivery process. Then there's all the people converting to propane for heat, water heat, generators...the demand has gone up too.

I don't think you'll damage those big cells with an open ended absorb time. Are they reaching the end-amps setting yet? They shouldn't be. Once the owners get comfortable taking sg samples often (daily at least, on a pilot cell or two) then they'll realize how long they need to absorb and at what voltage. If they're sulfated somewhat already then their internal resistance (i think) will show the voltages high while the SOC is low. That's the problem with using or relying on monitors until the SOC is known by SG reading and programmed in. They drift too. I don't ever rely on mine for SOC, just looking at voltage, amps in and such.

Best of luck with everything. And a happy customer will tell some people what a good job you've done, an unhappy customer will tell everyone what a terrible job they think you've done.

Ralph

Chris Olson

23rd February 2014, 10:34

Rob, I just browsed quickly over this thread and didn't read it in detail. I would've done a few things different for this installation. I would not have used the Magnum inverters. They have terrible voltage regulation, not exceedingly high efficiency (high temp transformers) and for a split-phase system the big XW6048 is king of the hill. Just the transformer in a XW weighs more than a whole Magnum inverter. The XW will handle massive overloads for quite some time (7.2 kVA for 30 minutes), and using generator support (which the Magnums do not even have) you could've saved the customer some money on batteries and generator, and long term operating costs of the system. You're going to find long term that the operating cost of the LPG generator is a killer for an off-grid system.

This is really a small system in my world. I can't even remember the last time we used less than 10 kWh in a day here. 20-30 kWh per day is more the norm for us in the winter time. And we run central A/C system here with our XW, completely off-grid. The design of the system to do that gets rather involved and I tried numerous schemes, including AC coupling a Jake 23-10, that didn't work. In the end I went with a diesel/RE hybrid system that uses a very small prime power diesel genset and the big XW synced with it on Generator Support. The diesel supplies base power for heavy continuous draw loads like central A/C @ 2.0 kVA on .17 gal/hr fuel burn. The inverter supplies the balance above that. The solar and wind can power all the loads during the day, unless the inverter overloads at which time it brings a much larger standby genset online to help out until the load drops below 6.0 kVA. At night we bring the small diesel online to take the load off batteries (and reduce the size requirement of the bank). We can power our grid-sized loads totally off-grid for less out-of-pocket expense than people in California pay for grid power. Like I said, it's a really long story, and quite involved, as to how I set our system up. But it works and it's proven, powering our off-grid home with central A/C all last summer without a hitch, and with minimal expense.

The other thing the Schneider system provides for an installer is the ability to monitor and make settings to the system from your desk at home instead of having to be onsite. This is done via the ComBox.

https://lh4.googleusercontent.com/-EC7-DnjuNgI/UwguIxI5jPI/AAAAAAAAH14/hU7-p5cbkD8/w905-h530-no/Screenshot_2014-02-21-22-54-43.png

As to the Surrette batteries and sulfation problems - having talked to Steve Higgins at Surrette quite extensively about this - it's again a really long story. But, the main problem is that even Surrette's light duty line (4000-series) has extremely heavy plates compared to other brands and they are more equivalent to Trojan's IND line of batteries. The charging specs for them are derived by the use of an industrial constant-current IUIa profile charger. Nobody in the RE business builds one. The "workaround" is to use higher absorb voltages. 2.50 VPC for absorb, and do not exit absorb based on a timer. Use 2% C for exit current from absorb stage.

If you cannot get the batteries to 2% C actual net current at 2.50 VPC within 180 minutes, then increase the absorb voltage more. We use 2.58 VPC on our Surrette 5000's here in the winter time. And we reduce that to 2.50 VPC in the summer time.

If you have Surrettes that are already sulfated from being deficit charged, and you live in the northern climates, you are going to have to burn a bunch of diesel fuel (or in this case LPG) to fix it. The longer you let them go, the worse they will get.

This is where the Schneider system would've been nice for you as an installer, Rob. You could start and stop their generator and make the appropriate settings to the system from the comfort of your easy chair in your office and take care of this for the customer. The customer is not going to do it properly in most cases.

Initially you'll have to set absorb to EQ voltage (2.60 VPC), disable the amp-based absorb exit, and use the timer set at 180 minutes. Don't bother with the hydrometer at first. Just watch net battery current during absorb @ 2.60 VPC for the next 10-20 cycles. For a 1200ah 48V bank you should be seeing <24 amps @ 2.60 VPC before the 180 minute timer expires. When you start achieving that <24 amp net current before the timer expires, then you'll have to get in the truck and drive to the customer's location to check SG's.

Set the absorb to 2.50 VPC, exit at 2% C, timer at 180 minutes. Do a normal absorb and let the system take care of itself. Check the SG's and you should have 1.255 or better (temp corrected). If you do not get 1.255 or better, increase the absorb voltage. The system will exit absorb based on either net current to the battery, or the timer. If it tends to exit on the 180 minute timer instead of net current, increase absorb voltage beyond 2.50 VPC. It will take about a month to bring a set of these heavy plate batteries back to "normal" using EQ voltage if you attack the problem every single day. Hopefully the RE will help out to reduce the gen fuel burn. But if it don't, don't let them go - burn the gen fuel instead and do it every single day. In the long term the gen fuel is way cheaper per kWh than those batteries (read above about a diesel/RE hybrid system saving battery costs in the long term).

For off-gridders generators and fuel and wind turbines and solar panels are not the most expensive thing. Batteries are. And the larger the off-grid system, the more impracticable batteries become. You still have to have them. But using RE power and diesel genset power to manage them by reducing the number of cycles, and the amount of amp-hour capacity you have to have, becomes pretty important.
--
Chris

Ralph Day

23rd February 2014, 10:59

Hooray! I was hoping Chris would pop in on this thread. The voice of experience.

Ralph

Chris Olson

23rd February 2014, 12:17

Hi Ralph,

I've been gone quite awhile. My wife and I have been sailing our yacht in the Caribbean most of the winter. When we came home to winter it was a shock. :eek:

Our yacht is slipped in Miami at present and we're leaving again in March for a 3-4 week cruise to someplace where there's not snow and ice. We left in December and we had house sitters here taking care of the place for all of December and January. Our system performed flawlessly for them and they couldn't even tell that they weren't on grid power until one day in December when there was an ice storm and everybody else's power was out. They went into town to get something and found the whole town locked and shut down because of no power. :)

Steve Higgins at Surrette used to work for Outback. Surrette's main business for years was industrial and marine. Jamie Surrette took over in the late 90's IIRC and today their RE business has grown to be almost half of all their sales. When Steve was hired one of his jobs he inherited is to re-write their battery manuals that presently reflect industrial charging profiles to reflect how RE chargers charge batteries with the old bulk/absorb/float algorithm. But for now he just tells folks to get that absorb voltage up. Few RE systems can constant-current charge a battery for 8 hours like an industrial charger does on grid power. So it requires different voltages to get the job done in the time you have to do it.

Rob Beckers

23rd February 2014, 16:37

Thank you much for the battery details Chris!
I'll set the system to use 62.4V - 180 minutes for absorb for the time being. Hopefully that, and a few more equalize opportunities at 63.5V, will bring the batteries in line again.

Regarding the choice of materials: In a world where money is no object I would have gone with the Xantrex too. They are however nearly twice the price of the Magnums. We, and the customer are actually quite happy with the Magnum system overall. It's working very well. The same goes for size; if you're paying an installer to put this in then 10kWh/day is not a cheap system, and definitely on the large end of the spectrum of systems I've been involved in. If you're doing the work yourself and labour cost is no object then the situation is of course very different.

Anything other than a propane generator was not an option either in this case. The customer is just not going to mess with diesel or gasoline. No matter how superior one of those solutions could have been. You have to play the cards you're dealt..

This system was designed with minimal generator use in mind. The amount of PV is sized to keep up with demand even in the middle of winter for our sun-hours and climate. Of course there will be periods in winter where the sun doesn't show enough, and the generator is needed to recharge the batteries. Other than those (hopefully few) times PV should be sufficient to keep the batteries charged day-to-day. Even now, in February, that is bearing out as the battery monitor (despite all its flaws) does show that the batteries receive the Ah's that were taken out, plus losses/efficiencies, every few days.

-RoB-

Chris Olson

23rd February 2014, 17:15

Rob, I think 63.5V is a little high for EQ. I'd limit that to 2.60 VPC (62.4V). Basically, the EQ is never recommended unless the cells are more than 20 points out of whack with one another. In this case, where I suspect the cells are simply sulfated from deficit charging, using the 2.58 VPC absorb should fix it. But don't expect to see dramatic results in just one or two times. It will take 20 cycles or so and you'll gradually see them start to come back (depends on how long they've been deficit charged).

Long term, using the higher absorb voltage has totally eliminated need for EQ's here. I get SG's between cells that are so dead even that it's never been needed after my initial panic when we first got the batteries and I discovered they were barely breaking 1.230 SG after what was supposed to be a "normal" absorb charge. These Surrettes seem less efficient at the onset due to the higher absorb voltages they take vs Trojan et al. But holy crap do they ever retain their capacity with age! Ours are four years next month and they still load test at 100% new capacity on a deep discharge test to 80% DoD.

And that's one more thing I would tell your customer about their new batteries - don't baby 'em. Work the snot out of them. Contrary to popular belief you can discharge those batteries to 80% DoD repeatedly and it won't hurt them. It actually helps keep them healthy by continually "cleaning" the plates and keeping fresh active plate material exposed to the electrolyte. Surrettes got real plates and they can take that for year after year and never erode the plates to where the battery is junk. If you would've bought them like T-105's or something, then it's different and >50% DoD repeatedly will severely shorten their life. But if you baby those Surrettes and only discharge them to 20% DoD you'll find they actually lose capacity within a couple months. Rolls even has a white paper on that and IIRC that paper says at least once every 30 days to deep discharge them to keep them happy.

Chris Olson

23rd February 2014, 19:12

Rob, I must say it's very impressive that you figure you can meet 10 kWh/day most times in the dead of winter with only 8 kW of installed solar capacity. We have expanded ours to 6.75 kW installed capacity over the last year. For our house sitters while we were gone, these are the graphs of our daily loads vs PV for December and January.

https://lh6.googleusercontent.com/-Sp9gDqa2ykk/UwqZidgnMfI/AAAAAAAAH2I/rdmj_utNsqo/w905-h530-no/Screenshot_2014-02-23-18-57-30.png

https://lh3.googleusercontent.com/-gS9_ssVcYqY/UwqZxJw9WYI/AAAAAAAAH2Q/kCp9sLpr7no/w905-h530-no/Screenshot_2014-02-23-18-58-28.png

With another 1.75 kW installed capacity we might be able to make 10 kWh a few more days in January. But I don't think December would be possible for us. I don't think even 30 kW installed capacity would do it in December for us, but for a very few days. Fortunately, wind works very good here, so we have tended to rely more on wind power in the winter than solar. Most folks I have talked to have never understood how that could be. Now that I got hard data from our ComBox that shows why, it is easier to understand why solar alone (for us) would never work, even at 10 kWh/day energy consumption, in the winter.

Rob Beckers

24th February 2014, 06:58

Chris, it's not magic, it's math and the PV statistics for our area. This won't work by looking at any particular year as your graph does, since there is huge variability in sun-hours from year to year. It does work over time though. I've assumed some pretty low efficiency numbers for batteries and inverters, and am comfortable with the 8kW in PV meeting demand "on average" in winter. If you want me to go over the number I'll be happy to do so, just don't have the time right now.

It's just too expensive to size the system such that even in a bad winter (as we had this year, with many, MANY snow days) it meets demand (nearly) all the time. It's sized to meet demand in an average winter. It was presented as such to the home owners: They were very well aware that in the middle of winter there are times that they need to run the generator (and that's one of the reasons we did install a generator).

This is working too: Since mid-January they have seen the battery bank bounce back to "full" just from PV every few days (we need to work on getting the absorb Voltage set high enough for these batteries, and to undo sulphation from running with low charge settings and user behavior but that's another matter).

Again, it comes down to cost. The budget is finite, we try to do the best with what we have.

-RoB-

Chris Olson

24th February 2014, 09:01

Chris, it's not magic, it's math and the PV statistics for our area. This won't work by looking at any particular year as your graph does, since there is huge variability in sun-hours from year to year.

Rob, this is what I find interesting about all this. Logging the system performance is a good tool. The ComBox just came out this past year so you have to start someplace with keeping track of it to make decisions on how to manage it. Most of this is seat-of-the-pants otherwise, with a lot of installers putting in systems based on theoretical solar insolation figures provided with tools like PV Watts. And then when the system doesn't perform as predicted because of localized conditions the customer is unhappy with it. Seen that many times in this area, primarily with folks who have grid-tied systems.

For our area we actually have had better solar insolation this year than normal. When Lake Superior freezes over (happens about once every 20 years and the last time was 1994) we get less snowfall than normal (240" in a season is normal here). This is combined with extremely cold clear weather with lots of wind from the north and northeast across hundreds of miles flat lake ice. Those are the conditions we're having this year - 30 to 40 below zero, weeks at a time when the temp never gets above zero (farenheit for us folks), and less snow from evaporation off the largest body of fresh water on earth.

So my graphs actually show pretty much ideal winter conditions for solar in this locale. The normal is much worse. I don't know what the latitude is of your off-grid installation there compared to ours. But in this area, folks up by Thunder Bay north of Lake Superior typically get much better solar insolation conditions in the winter than we get south of Superior. So I was simply comparing what we would have to do to try to power our system totally with solar and forget the headaches with maintaining wind turbines. I've looked at this for years and in the best conditions the turbines still reduce the cycling load on the battery. In the worst winter conditions those turbines are indispensible because the solar doesn't work at all for 6 weeks at a time, many times panels being buried under 3-4 feet of snow in 20 foot deep drifts. And that includes roof mounted panels. Ground mount is a non-starter here because even if the panels are clear the snow is so deep that the sun can't get to ground mount panels.

So that's what I'm looking at - if I take the values from PV Watts and compare them to my actual logged figures from this year, they compare pretty good. In the more normal year, PV Watts is grossly optimistic for this area. I don't know exactly how those numbers that PV Watts comes up with are derived. But I suspect it is a "model" based on latitude and theoretical hours of sunshine, and does not take into account localized conditions such as the largest inland sea in the world affecting the weather for thousands of square miles.

I know lots of off-grid folks up in the boonies NW of Thunder Bay that survive on solar panels alone (and lots of generator hours in the winter). I've always wished we could do that here, and using our prime power diesel was one of my theories on how it could be done with less turbines and the associated maintenance headaches with them. The diesel has proven less expensive to operate and maintain than wind turbines. However, the wind is still nice because it provides a 24/7 power source that reduces the cycling load on the battery long term.

Been struggling for years to get our off-grid power costs down below 50 cents/kWh, with equipment costs amortized over 10 years. The addition of the prime diesel finally got us below that, but it depends on winter time prices of off-road #2 diesel fuel.

Rob Beckers

24th February 2014, 15:31

For us this winter has had many more snow days (and no sun) than we usually get. So the opposite of what you're seeing. Snow cover of panels is a non-issue for this particular off-grid install; we have those panels at 60 degrees tilt in winter (30 in summer) and they are high enough that the snow piling up on the ground will not cover the bottom. For roof or other installs where snow is an issue I do actually take that into account in the numbers.

I do not use PV-Watts for off-grid, but run my own model based on sun-hours for every month of the year for this region. It is a very conservative model, and for most months or years the customer should actually come out ahead of what I predict. You can use PV-Watts too, but most definitely need to adjust the efficiency factor in there.

PV-Watts is preconfigured for grid-tie. It's pretty good at that, at least for our part of the world, I use it for that purpose, with slightly lower efficiency numbers put in. Over the last 5 years that we've been doing PV every single one of my customers has produced more each year than I predicted, which is exactly as planned: New panels do better, and I very much prefer to under-promise while over-delivering. As those panels age their long-term averages will be much more in line with my production predictions.

What I want to emphasize is that these energy production predictions are not voodoo! Nor is it "theoretical", it's based on measured sun-hour statistics for the area. What it does not do is tell you from day to day what to expect, it only works for statistical average weather, and of course each year will be different. It is as close as I can make it though for sizing systems and come up with equipment requirements that are realistic. Of course I can propose to double the PV (or add wind), but I wouldn't sell anything because my quotes would always be much higher than anyone else. I do win bids because I can explain to people where my numbers come from, that they are not just a stab in the dark (the "industry standard" for quoting, usually undersizing a system greatly so they can be the lowest quote).

By the way, regarding those sulphated Surrettes, I forgot to mention in my previous messages that the customer had a broken MidNite charge controller. So the first month (in the middle of winter, darkest time of year) they were running on only half the panels. That, the reluctance to run the generator, and the 'normal' (low) charge Voltage settings is what must have gotten those batteries to the state they are in.

-RoB-

Chris Olson

24th February 2014, 16:44

By the way, regarding those sulphated Surrettes, I forgot to mention in my previous messages that the customer had a broken MidNite charge controller. So the first month (in the middle of winter, darkest time of year) they were running on only half the panels. That, the reluctance to run the generator, and the 'normal' (low) charge Voltage settings is what must have gotten those batteries to the state they are in.

I'm sure the inoperative controller didn't help. But in my experience and lots of other folks that have mistakenly assumed the charging specs are for RE controllers, they are going to sulfate anyway. I've lost count of how many folks I know who have 4000-series Surrettes and didn't catch it fast enough because they never got the hydrometer out to check - or worse yet, don't even own one.

The constant-current IUIa profile that industrial chargers use is more efficient than the voltage-based charging that RE controllers use. Those industrial chargers are designed to efficiently charge industrial batteries from 20% SOC to 100% SOC before the next shift and they have a constant-current high voltage finish stage that RE chargers don't have. The constant-current finish stage takes the battery up to EQ voltage as part of the normal charge algorithm. But it doesn't hold the battery at EQ voltage like an EQ does in a RE controller. It's a constant current finish that starts when absorb is complete and hold the battery at 2% C current rate, gradually letting the voltage climb until it reaches finish voltage.

People mistakenly think that what is published for absorb for the industrial constant voltage stage is going to do the job - and it will IF you can absorb for like 6-8 hours. But few (if any) RE systems are able to do that. It's kind of a mess that Rolls got into with their business for years being primarily industrial and now gotten into RE but the manuals and published specs don't work with RE voltage-based charging. Steve Higgins is supposed to "fix" that and get new manuals out Real Soon Now. Or they might even have this info published someplace in their whitepapers on their website but nobody (including me) reads it :embarrassedt:

The real problem is that most people fail to use their hydrometer. If you check those SG's like you should be doing it becomes evident very early that you got a problem. And that can happen no matter what the brand name of the battery is. Rolls batteries just seem to be more pronounced with the problem of undercharging because of their extreme heavy plates they use and they have never messed with SG's like some of Trojan's batteries have done where some had low specific gravity electrolyte, and then some in the model had high specific gravity electrolyte, and you're left scratching your head trying to figure exactly what you got.

Rob Beckers

25th February 2014, 07:31

Rob Beckers

25th February 2014, 07:45

The homeowner did another equalize last Sunday, just using the sun / PV (since the batteries were nearly full from the day before). SG numbers are definitely moving up: The highest cell is now 2.260, lowest is 1.220, and most are around the 1.235 - 1.250 mark.

I am happy to see that the batteries are responding so well! With this much money in batteries I get very worried...

Next time I'm at the site I'll reprogram the controllers with the higher absorb Voltage (62.0 Volt / 200 minutes), and let the sun do its thing for a few weeks.

Chris, by the way, that 63.5 Volt for equalize is just about in the middle of the range Rolls/Surrette recommends. Yeah, that's from the much talked-about "battery care manual" that seems to have dubious value. If that manual is so much out of touch with reality for their batteries Surrette should rewrite this; I see Steve Higgins mentions the need for a rewrite over a year ago. With batteries being so expensive and the high risk for screw-up people deserve something that's accurate. Rolls really ought to make this a high priority.

-RoB-

Chris Olson

25th February 2014, 09:14

Chris, by the way, that 63.5 Volt for equalize is just about in the middle of the range Rolls/Surrette recommends. Yeah, that's from the much talked-about "battery care manual" that seems to have dubious value. If that manual is so much out of touch with reality for their batteries Surrette should rewrite this; I see Steve Higgins mentions the need for a rewrite over a year ago. With batteries being so expensive and the high risk for screw-up people deserve something that's accurate. Rolls really ought to make this a high priority.

Rob, yes I know. But that higher voltage can really push a lot of amps to a sulfated battery when it starts to come out of sulfation and the plates start to clean off. In my experience with that, never leave them unattended if you use that. If you have decent power available the possibility of a runaway is a good one. Using 2.58 VPC for absorb is pretty safe and it will take longer. But it can be done unattended over many days with little chance of a runaway meltdown in the event the batteries start to pull more amps as they de-sulfate. This is normally not a thing that you want to stand around and watch all the time because it's pretty boring. LOL!

They WILL come back with repeated absorptions at 2.58 VPC. In the winter I leave ours at 2.58 VPC for absorb and set it back to 2.50 VPC for summer when we have more hours that they can absorb longer. It has worked very well here.

I agree that Rolls should put this as top priority. Steve has a lot of irons in the fire. Rolls Battery is pretty small company with only about 40-50 employees. So I think sometimes the resources get stretched a little thin there with not enough hours in the day to get everything done. With today's internet resources, however, I don't see why they can't publish a whitepaper or bulletin that explains all this. In the old days Rolls used to have a complete section of their website dedicated to all their whitepapers and tech bulletins. That disappeared when they re-designed the website. It was one of the best sources of information on batteries that ever existed and it's a shame that it disappeared. Fortunately, over the years, I downloaded and saved a lot of those PDF's. But they're copyrighted and can't be re-distributed, and I respect that, so I save them for my own reference.

Rob Beckers

25th February 2014, 12:34

Rolls/Surrette still has some articles up at http://support.rollsbattery.com/support/solutions
Be sure to also click the "see other XX articles" lines, there's more than they show on this page.

Everything that charges on this system has a temperature sensor installed. They are either bolted to a battery terminal (the inverters) or stuck between two batteries (for the MidNite controllers). It is my understanding that high cell temperatures will shut down the charge controllers, making thermal runaway impossible. Since this keeps coming up I'll verify with MidNite.

-RoB-

Chris Olson

25th February 2014, 12:45

It is my understanding that high cell temperatures will shut down the charge controllers, making thermal runaway impossible. Since this keeps coming up I'll verify with MidNite.

Rob, it is possible MidNite has continued to improve firmware so it does that now. As far as I know it never shut the controller down with the firmware I am running. It just adjusts charge voltage per the temp up to a limit that you set in the controller.

But I could be wrong on that. I am not actually using the Classics anymore. We switched to XW-MPPT60-150's for solar. I am still using Classics with wind power, but wind power is typically too variable to actually charge batteries with it so I don't use the temp sensors on the wind Classics. It should be easy to test though - simply take the temp sensor off the battery and heat it up with your wife's hair dryer to see if it shuts the controller down.

Ralph Day

26th February 2014, 06:33

Now Chris, don't be sexist...you don't know just how much hair Rob might be hiding under that big hat:D. I use hair dryers for warming up model airplane coverings.

Ralph

Rob Beckers

26th February 2014, 07:02

A couple more of these jobs and I won't have any hair left... (Nor a hair dryer)...

-RoB-

Chris Olson

26th February 2014, 11:09

Ralph - I have to get me a hat like that. You can't hardly buy those anymore :D

RoB - aren't off-grid systems fun though? They keep you on your toes and I ain't never seen one yet that works out as planned out of the gate. They require constant tweaking, attention and maintenance. Grid-tie is so boring by comparison - it's just install and forget with an occasional lightning strike and a blown inverter to fix. :unsure:

Check out this off-grid system. It's in Zion Natl Park. There's some sections of the park where private homeowners are building huge elaborate off-grid homes there for some time now. And systems like this are the norm for these $5.5 million off-grid homes inside the park boundary. This system is 120/208 three-phase and the standby genset is a 150 kVA Cummins/Onan L-10 diesel.
http://youtu.be/gmSceBy39OY

Rob Beckers

27th February 2014, 07:03

Yeah, I saw that video of the extreme off-grid system. I'd like to see what they use for batteries for something this size. Fork-lift batteries?

I agree that off-grid systems are a whole lot more interesting vs. grid-tie. Though there is a difference when it's your own system and you don't have to drive half an hour each way to make a small setting change. I was out there yesterday again to change the charge controller settings; as discussed they are now at 62V and 200 minutes for absorb.

You were right Chris that the MidNite Classic doesn't check for battery temperature other than to change the Voltages. boB at MidNite though it was a good idea though, and added the code for it yesterday. So the next firmware release should have temperature limits where the controller will cut the current, to avoid thermal runaway of the batteries. You can't buy that kind of customer service!

-RoB-

Chris Olson

27th February 2014, 07:46

Yeah, I saw that video of the extreme off-grid system. I'd like to see what they use for batteries for something this size. Fork-lift batteries?

The batteries are 2V cells and if you watch this short video you can catch a few more glimpses of the system
http://youtu.be/BWwDYBr0gLw

One of the engineers I work with at Schneider told me about that system and I could ask her as she was involved with it, but it looks to me like they used 2V cells out of forklift batteries. They simply removed them from the forklift case and put them in custom cases. Forklift batteries have the type of terminals shown in that one snapshot there.

This is the home that the system powers, showing off the backlit marble floors:
http://youtu.be/8tXsg6xr5V0

People have varying reactions when they see something like this. Some say it's totally ridiculous and believe that if you live off-grid you should be living in a dark cave with candles and a single light bulb that you only use in an emergency. I've always firmly believed in the sky being the limit and I admire the engineering it takes to build a 66 kW off-grid system, and then show that it actually DOES work.

Chris Olson

7th March 2014, 17:43

RoB - sort of curious how your customer's batteries have responded to using higher absorb voltage?

Rob Beckers

13th March 2014, 08:00

Chris, I'm waiting to hear back from the customer with SG values. Since changing the absorb Voltage I've not been back there. We have had some good sun and very low temps since then, so I am hopeful that things are back to normal.

I will post when I get SG values.

-RoB-

Chris Olson

13th March 2014, 08:40

Cool, thanks. I'm hoping he/she is starting to see some gradual improvement. For us and a few other folks I know it has always been a slow thing - sort of an incremental increase in the SG with repeated cycling using the higher absorb voltage. If they have gotten back to 1.255-1.260 or better I would leave the settings alone and monitor it for awhile. If none of the cells ever go over 1.275 then they are not being overcharged.

With our Classics, in summer when the batteries tend to not get cycled as deep and get absorbed more often, I turned Absorb V down to 2.50 VPC to cut down on water use.

However, when we got the XW system with MPPT60's then everything is coordinated over Xanbus and the controllers have very accurate measurement of Net Amps going to the battery. The Classic has this now too with MidNite's new shunt dongle, but it didn't at that time. Using End Amps for the controllers to determine end of absorb with accurate measurement of Net Current to the battery allows me to leave the Absorb V at 2.58 VPC even in summer without overcharging or excessive water use.

Ralph Day

13th March 2014, 09:55

A little off topic but here's a quote I found this morning. The topic was about Florida woman being arrested and penalized for not being on city water/electric/sewer.

"i know in wisconsin you legally have to pipe your solar energy into the grid and the power company sends you a check for the differance. Its also illegal to collect rain water here. Or to use hommade windmills and such."

So Chris, is someone talking through their pants here? I kinda think so. It's amazing what people will say on the internet. Those homemade windmills...gotta keep people in line!

Ralph

link for story: http://www.collective-evolution.com/2014/03/09/florida-makes-off-grid-living-illegal-mandates-all-homes-must-be-connected-to-an-electricity-grid/

Chris Olson

13th March 2014, 10:20

We have to have an engineering stamp on towers for either wind or big solar arrays. But that's a safety issue to insure strength is adequate for wind loading, ice loading, etc.. And to insure that proper setbacks from property lines, overhead utility lines, buildings, roads, etc are adhered to so as to not present a hazard to persons or property.

But if we have to "pipe" our solar into the grid and the electric company is supposed to send us a check for it, I've never seen it. Don't even have grid to "pipe" it into, for pete's sake. And the Solar Panel Cops have never been here yet.

Nor has the DNR ever been here and said anything about our rain water drainage pond that our beef cattle drink out of. And we got an old Aermotor water pumper here that's been running for over 80 years and filling up a cistern for when our pond goes dry. Plus our electric wind turbines that can be seen on horizon up to 5 miles away. And the Windmill Cops have never been here yet.

And I don't expect to see any of these Solar Panel or Windmill Cops showing up any time soon. In fact, I'm pretty sure they could care less. There's at least two dozen fully off-grid farms and homes within 20 miles of us here. And there's no way they could keep utility lines standing thru the first winter here anyway. They wouldn't make any money on it even if they did run them because of the distances involved for the number of residences they could charge for it.

So let's just say that I pay little attention to the sky-is-falling media sensationalism that seems to be so common these days. They have to do something to get readers and viewers so they can sell advertising. And it seems to work for them for the most part because the general public sucks it right up.

Ralph Day

13th March 2014, 10:45

Ah Chris, I would dearly love to see a meeting of your local Homeowner's Association in action. Take this in the humorous spirit it is meant...trying to control a meeting of a roomful of Chris Olson type homesteaders would be like herding cats, big cats, a lot of big cats, with big claws and teeth. That's a meeting I'd like to see, if you could get them all in the room! :D

Some of the stories people have about HOA's are comical/frightening. Pushy busybodies with a lot of clout if you signed the HOA aggreement. I feel sorry for anyone living in such developments. I can't see my nearest neighbours, and we all get along just fine.

Ralph

Chris Olson

13th March 2014, 10:56

Ralph, there is not such a thing as Homeowner's Association here. The only meetings I know about are informal ones usually held at Jeff's Bar in town on Friday nights. And those are usually to plan deer hunting strategies, snowmobile or ATV trips, or simply to drink beer, eat pizza and play pool.

Ralph Day

13th March 2014, 11:07

I was just imagining...

Chris Olson

13th March 2014, 12:08

I was just imagining...

Well, imaging such a thing here wouldn't even work. The population here consists of two groups - Scandinavians and American Indians. Northern Wisconsin and Minnesota were settled long ago by Scandinavian immigrant farmers and loggers that tend to have rather independent and strong views on things. I'm only a second generation Norwegian-American and I don't even have an American wife. These virtues have been handed down over time and folks tend to take a very dim and suspicious view of socialist things like Homeowner's Associations.

Brian McGowan

26th March 2014, 23:59

I just read through this entire thread. It will take me a while to digest. A thought occurred to me while reading about the generator. Is it located in a position to take the heat it generates into the house to help with heating the house?
I was also wondering how much the battery pack cost. If I read correctly it is a 48 volt 1600Ah pack?

Rob Beckers

27th March 2014, 06:55

Brian, correct, the battery bank is 48V/1600Ah. The generator heat is not used for anything. The generator is not used much most of the year, just a few times in winter (and its set to run an exercise cycle every 10 days for a short time, if it hasn't already run for another reason).

-RoB-

Rob Beckers

6th April 2014, 19:26

RoB - sort of curious how your customer's batteries have responded to using higher absorb voltage?

Chris, to finally get back to this: It took a while, but as of today I think we have those batteries back to normal. The higher absorb Voltage alone was not cutting it; an equalize done a few back still had some cells lagging (1.23 after equalize for some). I told the home owner to equalize whenever the opportunity was there (ie. two days of good sun in a row so day 1 recharges the bank and day 2 gets them in float early in the day). A few more equalize cycles later and today's values have the cells as follows:

1.260 1.263 1.270 1.260 1.265 1.260 1.270 1.265 1.265 1.275 1.265 1.260 1.250 1.265 1.265 1.250 1.270 1.260 1.255 1.260 1.270 1.255 1.255 1.260

That's close enough for government work, so I've told the homeowner to equalize once a month from now, and if SG values prove to be within 0.02 he can go to once every two months.

By the way, I really like those HydroVolt SG meters! Super easy to use, no mess, automatic temperature compensation, very fast, very accurate. Can't imagine it would get any better!

I'll also swing by the place soon and change those absorb Voltages back to more regular values.

With the 8kW of PV sized for winter use, they are now swimming in electrical energy. Most days the bank is back to float by noon. I suppose it is a waste to not use all that energy the modules could be making the rest of the day, but such is life. Unlike you the average home owner just does not have the ability to tinker with their system to heat hot water and such. I'm happy if I can tech 'm just to take decent care of their batteries.

-RoB-

Brian McGowan

7th April 2014, 00:49

So there you sit in exactly the position I have been in so many times even though my system is so much smaller. This is the reason I say that this industry is lacking the needed equipment from an entire system perspective. They are focused on charging batteries or selling back to the grid. No thought is given to using energy that is generated beyond what is needed to charge the batteries. Once the batteries are charged they start choking down the panels. My charge controller provides a singe auxiliary relay that can be used for a low voltage disconnect or a dump load or to turn on a light at night. I haven't seen any others that provide more than that. I haven't seen any company that offers add on boxes for their controller that have additional controls for distributing excess power. You could buy add on boxes for your system but they would have their own voltage sense and I can tell you from personal experience in the QA and test equipment world that everything has it's own opinion of what the voltage really is. Even highly calibrated meters will have slight differences. This is why it is important for there to be one "brain" in charge of the entire system and nobody makes that now and it is not even something they are thinking about. As I stated in another thread there needs to be an add on box controlled by the charge controller that has a bank of 8 relays that can be programmed to be either low voltage disconnects or dump loads. As low voltage disconnects there needs to be a programmable time delay so that the voltage would have to be below a certain voltage for a given period of time so that it wouldn't shutdown when there was a momentary sag in voltage due to a heavy load kicking in. As dump loads one needs to be able to program them to come on in combination with each other to equal the amount of incoming power. There also needs to be analog inputs for current sensing so you can know if a load, such as a water heater, has been satisfied and is no longer drawing power so you can turn on another load to make up for it. As a dump load of last resort one should consider a hydrogen generator. I hear people all the time saying that hydrogen generators are not efficient but one has to consider whether generating hydrogen is less efficient than choking down the panels which is literally a complete waste of energy.
I know exactly how I would do all of this. I have contacted 2 charge controller companies but have gotten no response. I am trying not to make this myself but it is looking like I will have to because no one else sees the value in putting all the power that the panels or other sources are generating to use once the batteries are charged.
Brian

Chris Olson

7th April 2014, 11:29

1.260 1.263 1.270 1.260 1.265 1.260 1.270 1.265 1.265 1.275 1.265 1.260 1.250 1.265 1.265 1.250 1.270 1.260 1.255 1.260 1.270 1.255 1.255 1.260

I'll also swing by the place soon and change those absorb Voltages back to more regular values.

Those SG's are looking pretty nice with a few "laggers" in the 1.250 range.

My opinion is that you have a few cells that still aren't fully desulfated. I don't know if you're using ending amps on the Classics or just timers. If you're using end amps (the best) leave the Absorb V at least 2.50VPC and 2.58VPC is better. If you're using the timer only you can turn Absorb V down a bit but extend the time an hour until those "laggers" come around to your way of thinking.

Even using the high Absorb V I just finished with 11 hours of equalization charging this past weekend on ours to get the SG's back up to 1.265 after a winter of abuse. We had nice sun for two days and I used the little genset on Saturday morning to get them to Absorb V right way and get the EQ going. Our XW system refuses to equalize batteries until it finishes the absorb stage and gets them down to ending amps first. The Classics will just skip the absorb and go right to equalizing, and I think MidNite needs to fix that because it's not the right way to do it. But still, it took a little "boost" from the generator here to get them done absorbing early enough so I could use the peak solar thru the day to EQ.

I managed to get 5 hours in on Saturday and 6 hours yesterday. I got a few "laggers" that are still hanging around 1.250-1.255 but most are 1.265-1.270 now. Used 3.8 gallons of water in 11 hours of EQ @ 2.70VPC. It's a yearly thing, unfortunately. But once they get fully desulfated they're pretty good until next year and I don't EQ at all thru the summer because the battery temp gets too high doing that in hot weather. I prefer to leave the Absorb V set higher and take care of the EQ a little at a time every absorb cycle. That keeps the peak battery temp below 30-32C on a day to day basis in hot weather. Otherwise a EQ at high ambient temps will spike those Surrettes to 45+C and it takes 3 days to cool them back down.

Chris Olson

7th April 2014, 11:42

1.260 1.263 1.270 1.260 1.265 1.260 1.270 1.265 1.265 1.275 1.265 1.260 1.250 1.265 1.265 1.250 1.270 1.260 1.255 1.260 1.270 1.255 1.255 1.260

RoB - the reason I think you are not done with these batteries yet is because you still have 20 point spread on some SG's. 20 points is reason alone to equalize charge. That's why I say to keep at it with higher voltages for a bit until those "laggers" come around. No two batteries are ever the same. But 1.250 is still too low to consider a cell to be "healthy" IMO.

Rob Beckers

8th April 2014, 07:14

RoB - the reason I think you are not done with these batteries yet is because you still have 20 point spread on some SG's. 20 points is reason alone to equalize charge. That's why I say to keep at it with higher voltages for a bit until those "laggers" come around. No two batteries are ever the same. But 1.250 is still too low to consider a cell to be "healthy" IMO.

Surrette doesn't specify a spread for when they suggest equalization, Trojan does in their battery manual, and they consider cells to be "close enough" when the spread is less than 0.030, and all cells are above 1.235 when fully charged.

I agree that those cells at 1.250 need some work, but over the summer they will have an easy life, get fully charged almost every day, and there will be an equalization schedule in place. That should be enough to bring 'm around.

The MidNite charge controllers are currently set for 200 minutes in absorb at 62V, very nearly equalization Voltage(!). That was part of the "consistently overcharging" plan to gradually bring them back in line (and it probably helped). I will change them over to end-Amps, something around 20A should be OK for a bank this size.

-RoB-

Chris Olson

8th April 2014, 07:53

RoB - it seems to me without looking that Surrette says 20 points and you should equalize. At the high voltage, I would use 2% C/20, or 26 amps (assuming 1320ah?) for end amps and make sure you use the MidNite shunt to measure it accurately. Otherwise if there is loads on the inverter during charging it will be off.

That's what I use here is 2% C/20 for end amps, max 240 minutes - and I will continue to use 2.58VPC for now until I verify that my "laggers" have come around to my way of thinking. Then I'll turn it down to 2.50VPC for the rest of summer. And when the days get short again, I'll turn it back up to 2.58VPC.

Just to note that even with those high Absorb V's we still have problems with these batteries over winter. Not bad enough to require "corrective EQ" as Rolls calls it. But bad enough so it takes a couple good days of high voltage charging to bring them around once spring gets here.

Fortunately, somebody was on top of your batteries there and caught the problem before it becomes a real problem. Most people don't and they end up with a real problem. Amazingly, the Rolls batteries seem to be pretty efficient with high voltage Absorb V here. I have to use a 87% charge efficiency factor in my battery meter for them while Trojans and most others are happy with a 85% charge efficiency factor. For whatever reason (maybe the extremely heavy plates that Rolls uses?) they seem to hold their "surface charge" for a long time and deliver quite a bit of energy to the loads off that surface charge before they get down to what would be considered normal at-rest voltage after charging.

I was unhappy with our batteries when we first got them. After a few years of living with them now I'm extremely pleased with them. They are tough as nails and will take incredible abuse and still come back. I think they are THE premium off-grid batteries - even tougher than forklift batteries.

Chris Olson

8th April 2014, 08:08

RoB - what the deal with that hydrometer that you use? You mention that it temp corrects all by itself? Did you calibrate it when you got it?

I use a regular glass one with a float in it and a thermometer in the bottom. It was sold by Deka IIRC. I calibrated it when I got it by pouring some acid in a tube with ml graduations on it and weighed the acid on a gram scale. Mine was 5 points light on reading so I have to add 5 points to my readings. Just wondered how the off-the-shelf calibration was on that fancy one you got.....

It's interesting to note that Steve Higgins recommends using a refractometer instead of a hydrometer, and I don't know that I agree with that. Refractometers have to be calibrated too and are generally too fragile to be used in the field.

Rob Beckers

8th April 2014, 16:18

The hydrometer I use is called "HydroVolt" and it's made by a little Swiss company named Compaselect (http://www.solacity.com/Docs/HydroVolt.pdf) (they are actually manufactured in Germany). That company makes a variety of hydrometers for various liquids and purposes. If you click the link above it'll pull up a PDF that explains how it works, and what the measurement error sources are. It is a very smart design! Here is a picture:

http://prismsolar.co.uk/shop/images/cache/Dc%20Accessories/hydrohires.270.jpg

We import and sell these (as does MidNite too, by the way), and I gave this off-grid customer one so they can maintain their batteries. This is not an expensive hydrometer (we charge 30 bucks for it), it works remarkably well though, and as mentioned, it automatically does temperature compensation (it has two floats that expand/contract at the same rate, cancelling out each others temperature error).

It's accurate too: Easily readable within 0.005. Since we initially had some very low SG readings and I mistrusted the hydrometer, I calibrated mine. It was absolutely spot-on. Compaselect also sent me a calibrated hydrometer so I could verify if mine was off, and comparing the two again showed the exact same readings.

It can't break (unless you drive the car over it), there's no dripping mess, it's fast, it's accurate. What else can you ask of a hydrometer?

-RoB-

Chris Olson

8th April 2014, 16:53

That's pretty cool. I'm going to order one of those from MidNite. If it turns out to be easier to use than the glass one that will be good. I used to have a plastic one with a swinging needle in it and you couldn't break that one either - I tried to break it because that hydrometer was no good and you could get different readings on taking two samples in the same cell. Being it couldn't be busted or trusted I finally threw it in the garbage can in one piece.

So I was sort of curious about this device you have there. Never seen one like that before.

Ralph Day

9th April 2014, 06:37

If you do get one Chris, you might have to get a longer piece of tubing for suction. My Surrettes, when the electrolyte is at the lowest of 1/4 to 1/2 inch below filler bottom, it is too low to get a sample. I got some tubing at a local hardware store and made a sucker about 1/2 inch longer than the one provided.

I've used it, a good glass hydrometer and a refract type. The refract needed to be calibrated to 1.265 acid (distilled water made for too high readings...low actual sg). Don't use it anymore. I still prefer the glass unit, the Hydrovolt is good, but after 10 or so cells squeezing the bulb gives a cramp in the thumb/palm. Great for pilot/individual cells, but for doing the whole bank, not so much (for me anyway).

Ralph

Chris Olson

9th April 2014, 08:19

I got a box of battery electrolyte here that is weighed 1.270 and I'll calibrate the Hydrovolt against that when I get it. I got a longer tube off another junk hydrometer I can put on it if it needs a longer tube.

I'll give it a fair trial. It looks a lot smaller than the glass one, which is probably about 40-45cm long. And the glass one tends to drip when switching cells so I always have to have a rag handy to wipe up any drops on the battery tops. Once calibrated, there is no chance of a glass one ever giving false readings. Any time you introduce mechanical bits into a measurement instrument there is a chance for error. So we'll see what it looks like when it gets here. It's supposed to be shipped today so I'll probably have it by the end of the week.

Rob Beckers

12th April 2014, 07:32

I would use 2% C/20, or 26 amps (assuming 1320ah?) for end amps and make sure you use the MidNite shunt to measure it accurately. Otherwise if there is loads on the inverter during charging it will be off.

We don't have MidNite's shunt, but I believe using end-Amps is still useful: To keep the Voltage at the absorb level current needs to be going into the batteries. Any net-load and the Voltage will go much lower. So, worst case the charge controllers are shouldering the house load with little going into the batteries during absorb, meaning that they would get (mildly) overcharged before the controllers decide they are done. Or the current never dips low enough to kick out of absorb, and the batteries get a trickle all day long. Neither scenario is a bad thing for the batteries.

The customer doesn't have your setup where excess PV energy is put to good use, so we might as well put it towards additional absorb time. According to Rolls' own manual these batteries are supposed to absorb for 5 hours (!!!) at the charge currents we can realistically supply through PV.

-RoB-

Chris Olson

12th April 2014, 09:32

RoB - actually the absorb recommended time might be even longer than 5 hours at normal voltages and it depends on DoD. It seems to me I have one of their charts from the old tech bulletins that explain that and ours figured out to like 8 hours @ 2.45 VPC if the batteries are discharged to 80% DoD!

We do pull our batteries down close to that far now and then and they seem to like it. But it takes us two days to get them back to 100% SOC again, tracked by our Trimetric. After they've been at 70% DoD it takes a perfect day of charging to get them back to 75% SOC on the first day. A second day of perfect charging to get them back to 100% SOC according to the Trimetric. And a third day to get the gravities back up to normal. Big batteries are fine, but if you don't have the power the charge them, they can be a liability instead of an asset.

I got my Hydrovolt and tried it out. It calibrated right on. I sampled some random cells with it to see how it works and I must say (for me) it is easier to use than the glass one. You don't have to sit there and bing it to get the bubbles off the float (although that "habit" has become automatic with me over time). Checking gravities on batteries is not really wildly fun in the first place, so a tool that makes it less messy and easier to do is more likely to get used. So I like the Hydrovolt. Ralph had mentioned the difficulty of squeezing the bulb, but it doesn't seem overly excessive to me.

Rob Beckers

12th April 2014, 16:33

Chris, same experience here regarding bringing the cells back after a deep discharge. We looked at SG after such an event, and though the cells would go through their bulk-absorb cycle the SGs were still well below "full". Must be similar to the Peukert effect, where the amount of charge taken up actually depends on how far the discharge was.

Glad to hear you like the HydroVolt!
I think part of how well one likes (or doesn't like) sqeezing that bulb is simply a matter of physiology: From your picture you look to be about the same build as me (I'm of classic Dutch stock, 6'7" with hands the size of small shovels, for better or for worse). Big hands make this easier. I can fully understand though where Ralph is coming from. May depend on how many cells you have to do as well: So far I've only used this (at most) on 24 cells at a time.

-RoB-

Chris Olson

12th April 2014, 16:59

Must be similar to the Peukert effect, where the amount of charge taken up actually depends on how far the discharge was.

I'm pretty sure that's what it is. The specs they give for discharging batteries to 80% DoD are developed in the industrial world with constant-current IUIa profile chargers and they are designed to get a fully discharged battery back to 100% SOC from the end of the shift to the next day. The IUI profile is much more efficient than the voltage-based charging we have in the RE world. And it would be darned near impossible to do IUI constant-current charging with solar power.

These are the things that Steve Higgins has promised to address in re-writing Rolls' battery manuals for batteries used in solar off-grid applications.

On the Hydrovolt, I'm thoroughly impressed with it. I went thru our entire bank today with it and measured the SG's and recorded them in the battery logbook. We got 72 cells here with three strings and it is so much simpler and quicker to use that it took me half the time. And no drops on the battery tops that I had to clean up. Every once in awhile somebody comes up with something that actually works, and the Hydrovolt impresses me as that sort of a thing. Thanks for pointing it out!

Chris Olson

17th April 2014, 12:47

Well, I'm happy to report that we got our Surrette Laggers back up to 1.260 finally. They sure were stubborn!

I looked in our TriMetric this morning at the cycle history and it shows pretty horrible cycle efficiency for the last five cycles. But now that we "got 'em back" I can turn the Absorb V down a little and try to get the average amp loss back to ~.40-.48 A over the length of the cycle. We've been running in the low to mid 3's for most of this month on amp-loss. That's a lot of power lost just to get the batteries back to 100%!

When somebody finally invents something that beats the venerable flooded battery for reliability and efficiency at the same price, you can rest assured I'll be waiting in line to buy it.

Rob Beckers

17th April 2014, 13:05

Chris, what do the Amp numbers for cycle efficiency mean? I've never used a TriMetric (heard lots of good things about 'm but never had the occasion).

-RoB-

Brian McGowan

18th April 2014, 01:18

What you gentlemen are speaking of here is exactly the reason I had to give the nickel iron batteries a try. I was spending a lot of energy just trying to keep my FLAs from dying before their time.
I would be interested in figuring out how I can tell the efficiency of my batteries so I can compare them to the batteries you guys are using. Any help would be appreciated.

Ralph Day

18th April 2014, 07:04

Chris,
I think I recall that you have the 2v surrettes. I have a couple of lagging cells in my 8v ones, one particularly slow, and \I charge those cells a little extra.

With the help of a friend I designed and built a 2amp nominal charger that I run on the affected cells to give about 2 amps extra current to just the one (or two if it's accross both cells). In my case I had to drill a hole in the plastic case just above where the 2v cells are bolted together, thread down a flattened bottom lag bolt until it touches, then clip on the charger leads. True, it leaves pos and neg terminal connections exposed and relatively close together, but I'm careful around them...no metal near (about 4-5 inches apart).

When the system is absorbing or floating I can turn that little charger on and it pumps a little more current into just 1 cell. (I also can plug it into utility power overnight). You would not need to cut into your battery cases. With the absorb times you run maybe the extra current would help the laggers without having to over charge the rest of the bank (that was my thinking with my setup).

I thought it was a good use of extra power during absorb and float, besides running my distiller for battery water, or a resistance heater.

Ralph

Mark Ross

20th April 2014, 21:57

1) If it was up to me, and I had to design a house to be off-grid from the get-go, I would have opted for a design that utilizes passive solar for some of its heating, and use a wood stove as the main heat source. Maybe a regular (propane) furnace would be part of the mix but I would try to do without having to use it for most days. Even though the furnace runs on propane, the blower motor is a very large load!

Any experience with "outdoor furnaces"? Notice there are some that burn both wood and come with duel-fuel options so they can fall back to propane. They provide hot water to the house which is suitable to heat domestic hot water and also radiant in-floor heating systems. Just recently discovered these as a possibility and wondered if you'd come across them before.

Thanks!

Rob Beckers

21st April 2014, 15:46

Someone I know down the road from me has an outdoor wood furnace; they are about the size of a (big) outhouse, and use insulated water lines running underground to get the heat to the house. I dug the trench for those lines with a backhoe that I have for this particular one. The unit is about 80 feet from the house.

The owner is very happy with it, seems it works like a charm and heats the house fine. You do have to go outside to fill it up, but they take a huge amount of wood (and burn just about anything), so you don't have to do that too often each day. The thermostat controls a fan in the furnace, it 'simmers' when no heat is called for and only really burns wood when there's a need for heat.

As far as I know he has a regular air-handler that puts that heat into hot-air for distribution around the house.

-RoB-

Mark Ross

22nd April 2014, 17:06

As far as I know he has a regular air-handler that puts that heat into hot-air for distribution around the house.

-RoB-

Interesting, in that case I guess he doesn't realize the advantage of running the blower motor to move the air. In our case I am considering using an outdoor furnace and using that to heat a radiant in-floor system - unfortunately much like you trying to get quotes on doing radiant in slab is absurdly expensive, it just doesn't even make sense. May have to pick up that book you mentioned and have a crack myself.

Thanks!

Chris Olson

30th April 2014, 01:03

Chris, what do the Amp numbers for cycle efficiency mean? I've never used a TriMetric (heard lots of good things about 'm but never had the occasion).

RoB, in the TriMetric's history it keeps track of the last five cycles. Just a rundown on what it keeps track of (from memory without looking):
H1 = cumulative amp-hours
H2 = hours since last cycle ended
H3 = number of hours in last cycle
H4 = amp loss of last cycle
H5 = lowest SOC% of the last cycle

And several other things as well. The amp loss is the amps lost over the hours that the cycle ran. I'll include an example from our last cycle (as the TriMetric is a VERY useful tool for managing your batteries):

It says the last cycle lasted 74.2 hours. This is the time from full recharge to the next full recharge. Our amp loss was .17 for 74.2 hours = 12.6 amp-hours lost. This means it took 12.6 amp-hours more to charge the battery back to 100% than we got out of it. The lowest SOC was 81% during the cycle. With a 1200ah battery, that means our max discharge was 228ah. And it took 240.6ah to recharge.

Now, keep in mind this is on a system that uses extensive use of wind power. Wind power runs 24 hours a day. So in reality what happened is that the battery stayed around 90% SOC for most of the cycle due to the wind turbines constantly feeding the system. The draw down to 81% SOC was a couple days ago when Kristin did laundry and ran the electric clothes dryer for several loads. The day after that (yesterday) we got 39.7 kWh from wind and only 2.0 kWh from solar because it rained all day. The wind turbines got the battery up to 94% by this morning. The wind continued to blow, plus it continued to rain, and the battery got back to 100% by late afternoon.

Due to the battery being constantly trickle charged by wind power the cycle efficiency is quite good at 94.7% on that cycle. It was so efficient it did not raise the battery temp even 1 deg C during the entire cycle.

We typically cycle our battery on 7-10 days between full charges (the TriMetric tells you that too). So we can achieve cycle efficiency in the mid-90's most of the time. When the battery gets cycled down to 80% SOC, then immediately recharged the next day the cycle efficiency is horrible - usually at below 85%. If the battery is cycled down to 50% then fully recharged the cycle efficiency is better at usually around 90%. Cycling your battery only about 50 times per year results in the 93-95% cycle efficiency I have designed our system for.

Winter time we can go sometimes 2 weeks or more to full charges. This last winter we were gone and the battery didn't get fully charged for six weeks. The resultant cycle efficiency to equalize and get them desulfated again is very very low.

Mark Ross

30th April 2014, 20:21

Completely unrelated, and possibly real dumb question, but in the system described in the first post, there are two stacked Magnums, and also two Midnite Solar Classic charge controllers...but, the magnum's have chargers built in? Or are the magnum charge controllers only used for charging from generator, and the midnites are used to charge from PV array?

Chris Olson

1st May 2014, 00:09

The inverter chargers only use generator power for charging. The Classics use the solar panels for charging.

Mark Ross

1st May 2014, 08:17

The inverter chargers only use generator power for charging. The Classics use the solar panels for charging.

Thanks! Somehow I had never realized this until now. Learn something new every day!

Chris Olson

1st May 2014, 08:52

Thanks! Somehow I had never realized this until now. Learn something new every day!

Mark, I'm not sure of the design of the Magnum inverters but I think they only have one AC input. You can feed that AC input with either generator or utility power. The AC IN in a Magnum passes thru to loads on the AC LOAD side of the system and the balance is used to charge the batteries.

Different inverters do this differently internally. This is the inside of our XW inverter:

https://lh5.googleusercontent.com/f4Ovc9p92kONaaBZR5wKXKS2hPbw3KC6YsqHQbVP7Pg=w640-h480

The DC section of the inverter is on the right (with the exception of the main processor board which is on the upper right) and the AC section on the left. There's a big toroid transformer behind the DC section that weighs about 65 lbs alone. The XW has dual AC inputs - AC1 for grid, AC2 for generator. In our case, where we do not have utility power, we have dual generators on AC1 and AC2.

The XW, unlike a Magnum, synchronizes its sine waves with the AC input power source. With the inverter's output being in sync with the input, it can reverse directions of the inverting section to use the AC input source to charge batteries - or if the load exceeds the capability of the AC input source, it can use battery or RE power from the DC side to assist the AC input source.

Most inverters are not as sophisticated as the XW and they do not operate in sync with the generator. With these cheaper inverters that do not sync and just use a simple internal transfer switch, when the generator accepts load your loads can be "hit" out-of-phase when the generator accepts load. For resistive loads, not a big deal. For inductive loads like motors this is not good.

You get what you pay for when you buy an inverter. If the inverter is cheap, there's usually a reason for it.

Mark Ross

1st May 2014, 16:16

Mark, I'm not sure of the design of the Magnum inverters but I think they only have one AC input. You can feed that AC input with either generator or utility power. The AC IN in a Magnum passes thru to loads on the AC LOAD side of the system and the balance is used to charge the batteries.

Different inverters do this differently internally. This is the inside of our XW inverter:

The DC section of the inverter is on the right (with the exception of the main processor board which is on the upper right) and the AC section on the left. There's a big toroid transformer behind the DC section that weighs about 65 lbs alone. The XW has dual AC inputs - AC1 for grid, AC2 for generator. In our case, where we do not have utility power, we have dual generators on AC1 and AC2.

The XW, unlike a Magnum, synchronizes its sine waves with the AC input power source. With the inverter's output being in sync with the input, it can reverse directions of the inverting section to use the AC input source to charge batteries - or if the load exceeds the capability of the AC input source, it can use battery or RE power from the DC side to assist the AC input source.

Most inverters are not as sophisticated as the XW and they do not operate in sync with the generator. With these cheaper inverters that do not sync and just use a simple internal transfer switch, when the generator accepts load your loads can be "hit" out-of-phase when the generator accepts load. For resistive loads, not a big deal. For inductive loads like motors this is not good.

You get what you pay for when you buy an inverter. If the inverter is cheap, there's usually a reason for it.

Yeah, I really like the idea of generator support for sure. Right now I'm considering either the dual Magnums similar to what is described in this thread, or the XW6048.

I'm still not sure I understand how it allows for a smaller battery bank? (the smaller inverter part makes sense for sure)

Chris Olson

1st May 2014, 18:02

The XW6048 will be more expensive than dual Magnums by the the time you buy everything you need for it (SCP, conduit box, distribution panel, AGS, etc..). However, the XW6048 is more electrically efficient (by a fairly large margin). The XW6048 has much better voltage regulation - one of our off-grid neighbors bought a Magnum PAE and they couldn't even run their washing machine with it without the lights in the house flickering like a disco. After spending hours on the phone with Magnum and no results, they finally put in a XW4548 instead and sold the Magnum on craigslist. I helped them put in their new XW when they got it.

And the biggie - the XW has generator support and auto leg balancing for your generator. The Magnum doesn't.

The generator support does a number of things.
#1. It allows you to use a smaller more fuel efficient generator without fear of overloading it or having the inverter "spit it off" due the generator dropping out of voltage and frequency spec on overload. Without generator support if you need to run a 5 kW load you need minimum 6 kW generator to have a little reserve because the gen is pure passthru to the AC loads.

With generator support you can run that 5 kW load with a smaller generator. The advantage of a smaller generator is clear - generators are most efficient when operated at 80% rated load or better, continuous. It is virtually impossible to keep a 6 kW or larger generator loaded to that level on an off-grid system when it is running. We had a 6 kVA genset originally and it was way too big. Couldn't even afford to stick the key in the switch because it sucked OPEC dry in the matter of a few hours. We downsized to 4 kVA and got much better results.

Everybody that lives off-grid needs to use the generator at some point. Most people avoid it like the plague because of the cost, and because they got WAY too big of a generator. Read the first post in this thread - the people that have this system avoided running the generator in the winter when they should have, and look what they ended up with - sulfated batteries. You do yourself no favors by oversizing the genset. But with the type of system they got, they have no choice. They have to have the big generator because on pass-thru the Magnum will spit a little one off on something as simple as a surge load from the 'fridge compressor starting while it's charging batteries.

#2. You can use a smaller battery bank with generator support. Batteries are the most expensive consumable component of off-grid power. Not solar panels, generators, inverters, or wind turbines - batteries. You will spend more on batteries - JUST on batteries - for an off-grid system than you can buy the equivalent in utility power for, compared to how much energy the batteries can store over their useful life.

How does this work? If you need to run occasional loads over the capacity of the inverter you got two choices - either stack inverters or use a generator. If you go with the stacked inverter configuration, you need to double your battery bank. Read above on why you don't want to double your battery bank.

Even a big battery bank cannot power 8 kW worth of inverter output for more than a couple hours at full rated load. The Peukert Effect kills fully charged batteries deader than a rock in short order. With generator support you can power your 8 kW load for hours on end with a smaller generator and using the balance from your inverter and batteries. I put the pencil to it and the cost/kWh to use generator support for peak load is less than half of JUST what it costs for batteries to run a second inverter.

I already been the stacked inverter route in the past - and it was not pretty when it came to actually being able to use them. The voltage sag on the battery bank is so severe with two 4 kVA inverters at full rated load that you might as well forget it. When the voltage sags under load, it requires LOTS more amps - so your amp-hours in your battery disappear before before your very eyes if you try it.

Sure, you can power them during the day when the batteries have decent SOC and you got lots of incoming solar power. But what happens at night when the batteries are down to 80% SOC and your wife wants to do laundry or make a pizza in the oven? You going to tell her, "no, you can't do that - you have to schedule those loads for tomorrow when the sun is shining"? That goes over here about as well as the proverbial lead balloon. My wife just throws the switches, the system handles it by starting the generator and uses the generator as long as its needed - 24 hours a day, no matter what the weather is, we got 8+ kW on tap without having to interact with the system or even think about it - it "just works". You are not going to have that reliability and convenience with stacked inverters because you're always going to have to be checking the battery condition to make a decision on whether or not you got enough stored juice to power them.

The final thing is that the XW has massive overload capability. Ours will handle 7.2 kVA for 30 minutes before it starts flashing the red light on the SCP for overload. It will do 8.0 kVA for about five minutes. 12.0 kVA for 15 seconds.

The Magnum inverter line was originally designed for the RV market, and I think they pretty much dominate in the RV business. That's why they have a small form factor and an inefficient high temp transformer. They started selling them for RE to expand their market. But from experience I have had with them they are pretty light duty for full-time off-grid use. Just the transformer even in a XW4024 weighs more than a whole Magnum inverter. The thermal mass in the XW is what gives it its legendary overload capacity with no temperature based de-rating.

Our XW6048 is the only single inverter we have ever had that can power my Lincoln 225A spark welder straight up without a generator. Even our stacked SW Plus 4024's wouldn't power it because the stacked configuration can't come up on full surge capacity fast enough when you strike the arc. We had a SW Plus 5548 for awhile with a Outback PSX-240 transformer on it for split-phase. That one would do it with the generator pre-started, but the severe voltage sag when the arc was started caused the lights in the house to go dim and our HDTV would konk out. The XW does it without even a fuzz line appearing on the HDTV screen - although my wife reports she's seen our TriMetric go to 400A to the inverter when I'm welding out in the shop, which is probably around 20 kW of input power to that inverter on full surge - and it takes it - time after time without skipping a single beat.

That's the difference between a cheap inverter and a good one.

Mark Ross

1st May 2014, 19:40

#2. You can use a smaller battery bank with generator support. Batteries are the most expensive consumable component of off-grid power. Not solar panels, generators, inverters, or wind turbines - batteries. You will spend more on batteries - JUST on batteries - for an off-grid system than you can buy the equivalent in utility power for, compared to how much energy the batteries can store over their useful life.

How does this work? If you need to run occasional loads over the capacity of the inverter you got two choices - either stack inverters or use a generator. If you go with the stacked inverter configuration, you need to double your battery bank. Read above on why you don't want to double your battery bank.

Wow Chris, thanks for all that info. A lot to take in, in one sitting! So if I'm reading #2 right, if you use a single inverter you can have a 24V bank instead of the 48V? Is this the "doubling" you are referring to? Sorry, these must seem like pretty trivial questions!

I definitely agree, this system has to be "wife friendly"!

Chris Olson

1st May 2014, 22:54

So if I'm reading #2 right, if you use a single inverter you can have a 24V bank instead of the 48V? Is this the "doubling" you are referring to?

Not necessarily. Say you have eight 6V 400ah batteries. All in series, these batteries are 400ah @ 48V. The same batteries in series/parallel for 24V are 800ah. But the kWh capacity of the battery bank is the same.

When I talk about doubling the bank for dual inverters, I'm referring to the fact that you need to double the kWh capacity of the battery bank to handle dual inverters at full load running on battery power. Living off-grid, the times you tend to use the most power are in the evenings and winter time. You're working during the day if you're normal. So you're not around to take advantage of being able to schedule heavy loads during the day when you got the solar power to run them. Everybody gets home in the evening after work and you start doing laundry, and cooking supper and watching TV, etc.. - the loads go right up.

Due to this fact, stacked inverters might look pretty and awesome on the wall. But when it comes to actually being able use their stacked power output capability, they have to be fed. It's that feeding part that gets really expensive because batteries suffer from what is called Peukert Effect. Meaning they have less than nameplate amp-hour capacity if you discharge them faster than their 20hr rate of discharge. Therefore, feeding stacked inverters in an off-grid situation becomes an exercise in "how much money you got"? You can throw big wads of battery capacity at the problem, but then in the winter time you find out you don't have near enough solar capacity to charge the batteries, so you end up running the generator anyway.

Well, that's easy to fix. Instead of using the generator to charge batteries, we use generators to manage battery capacity. For heavy load situations, instead of sacking the battery out we use generator support to reduce the load on the battery to more normal levels (around the 20 hour rate of discharge). During periods of poor solar and wind conditions in winter instead of sacking the batteries out then charging them with a generator, we use a little tiny generator on generator support to reduce the discharge rate and make their capacity last longer until the sun and wind comes back. Then let the sun and wind do the charging.

We have a very sophisticated monitoring and logging setup on our power system. It logs every single watt-hour even from the generators and I have used it fine tune our generator support system. By achieving less than half the cycles on our batteries per year that conventionally designed off-grid systems do, and reducing how much battery capacity we need to support our 25 kWh/day lifestyle, we have our out-of-pocket expense for batteries (remember they eventually have to be replaced) and fuel at about 33 cents/kWh. That might not seem too impressive when you can buy utility power for 12 cents. But for off-grid power it's pretty good because 50-60 cents/kWh is more the norm.

Most people don't realize this, but even a small battery bank costs $2,000. If you use 6 kWh/day in your off-grid home by using propane for cooking, heating, etc.. - so all the electricity is used for is lighting and the basics - if the batteries last 7 years they cost you 13 cents/kWh. That's more than what utility power costs.

Living off-grid is very, VERY expensive. Been doing it for 14 years. I've learned a few tricks along the way, and one of those is that when you run a generator you'd better be powering loads with it direct. Changing it's power to DC to charge batteries at only 85% efficiency, then turning around and changing its power you put in the batteries back to AC to run your loads is pretty much an exercise in futility. It would be cheaper to just move to town and burn money in your BBQ grille for fun.

Mark Ross

2nd May 2014, 08:35

Thanks for all the info Chris, this makes sense. Our biggest issue is that our house isn't built yet so we don't know what all the loads will be, so makes it tough to plan out a system - right now just sort of shooting to over-engineer something a bit to give us some leeway. It's going to be a small well-insulated house with only my wife and I there for a majority of the time so hopefully we will be fine :wondering:

The real mystery loads are like the well pump and other systems I've never interacted with before...

Chris Olson

2nd May 2014, 09:12

The well pump don't really take much power unless you plan on irrigating your garden or something with it. It's intermittent draw and if you put in a decent pressure tank the starting and stopping for two people is pretty minor.

Other things like central A/C can be major power consumers for off-grid. We have central A/C in our place - a 1.5 ton unit. When it's running it uses more energy than all our other stuff combined. We use solar power during the day to run it, and diesel prime power at night in hot weather. There is no way you can power A/C all night on batteries. So to be comfortable and be able to sleep when it's hot, our little Robin diesel generator supplies the power @ .17 gallon/hr with the generator support level set at 2.2 kVA. The little Robin diesel can't start the compressor in the A/C unit by itself because the compressor draws 27 amps @ 240 volt for starting and 59 amps at locked rotor (with pressure on the head). But with XW assisting it it has no problems with it.

A/C compressors are a bad one. IF you are on genset power and the compressor attempts to start but the genset can't handle it so the generator gets "spit off", the compressor now has pressure on the head for an attempted restart. So you're now dealing with locked rotor amps. Central A/C off-grid is where generator support really shines because the system can use the surge capacity of both the inverter and genset for starting the compressor as it cycles. Inverters have way more surge capacity than gensets do.

So you need to get together a list of what you plan on. When you look at this, even a 4.0 kVA inverter running at 50% rated load 24 hours a day can supply 48 kWh to loads. The problem is peak loads. Come up with the worst case scenario for peak loads, then put the pencil to it and compare what it takes for equipment to power those peak loads with inverters and batteries vs using a genset on Gen Support. I'll think you'll find the same conclusion I came to after several years of trying to figure this all out - extra inverters and batteries doesn't make sense when you can buy a small diesel genset for the same price as an extra inverter, and the fuel and maintenance for that genset is only a fraction of the cost of batteries to power an extra inverter.

Mark Ross

2nd May 2014, 09:22

It sounds like generator support is the way to go for sure.

We won't have A/C so that won't be an issue for us - in the most brutal days of summer we can hopefully simply retreat to our cottage on the St Lawrence which is typically cooler (and has grid power for A/C!).

We'll also be using gas for the range (we don't have the same objections to gas/propane as you) so this will also be another load we don't have to worry about (as much, I understand the glow bar still draws a reasonable amount of power) and probably also leverage that propane for some backup home heating as well (but primarily heating via wood stove).

What size generator are you running on your system? One of the things I had perhaps mistakenly understood, was that the generator size depends on battery bank size? That you need to be able to push enough amps into the bank to charge it in a timely fashion?

cheers!

Chris Olson

2nd May 2014, 10:09

Our off-grid system is what Schneider Electric calls a "solar/diesel hybrid" off-grid system. I had years of experience with diesel gensets working as a mechanical engineer for Cummins Power Generation - mostly 1.5-2.5 MW class generators on prime and standby applications - but the principles are the same applied to smaller systems.

The basic premise is that no matter what you do you are not going to eliminate the use of fossil fuels for off-grid power. The nature of solar and wind is that it is unpredictable, and the methods used to store solar and wind power for later use (batteries) are expensive. So you use the renewable sources of energy to operate your diesel generator more efficiently and reduce how much fuel you use in it. That means you don't use it for battery charging.

Conventional design calls for sizing a genset on a battery based inverter system to be able to charge the battery at C/10 while carrying pass-thru loads at the same time. With a hybrid design where the generator is managed differently, the generator is sized to be able to carry peak loads at the desired battery discharge rate during peak load operation. You are managing battery discharge rate with your generator so you don't have to use it for battery charging later.

I do not work for Schneider but I have worked closely with their engineering dept on developing this management method. Incidentally, Sandia also developed a similar system for the USCG for their remote transmitter sites, and patented it. The Sandia system uses a DC generator because all the loads at the remote transmitter sites are DC. The Sandia system is call ACONF - you can look it up. Of course, Sandia spent taxpayer dollars on their system and produced a white paper on it outlining their pilot study when they ran a conventionally designed system alongside ACONF. They got 38% reduction in generator run time and 40% reduction in battery cycles compared to a conventional off-grid system. We have achieved much better than that here using a different generator management system on AC loads. And I'm working on a more advanced genset management system yet that will use measured SOC to determine how much kWh to inject into the system to control battery discharge rate. An overview of the ACONF system is here:
http://www.sandia.gov/ess/docs/pr_conferences/2007/corey_snl.pdf

On our system we use three different gensets. One is a 3.6 kVA prime/4.0 kVA standby unit used for peak load. The other two are smaller prime units - one is a 2.3 kVA gasoline fueled set, the other a 3.0 kVA diesel set. We only have two AC inputs on the inverter, but I have designed and built an auto transfer switch in the generator power room that selects which generator to use when the system calls for gen power, starts that genset, and switches to it on the AC2 input.

Mark Ross

4th May 2014, 18:41

What do you do when you simply have to recharge the batteries from generator? There isn't any wind around here and I don't think I can guarantee enough solar even if I oversize the # of panels.

Can you have two generators connected to the inverter, a small for gen support and a large for bank charging if necessary?

EDIT: Thanks again for all the info, sorry must be brief the internet is questionable up here at the property!

Chris Olson

4th May 2014, 18:51

Andy Rhody

4th May 2014, 20:50

Chris, if I recall you have both wind and solar. Could you update us on you set-up. Didn't you design your windmill, or windmills? How about some photos?

Chris Olson

4th May 2014, 23:03

Andy, yes the majority of our power (up to 80%) comes from wind in the winter time. And yes, I build the turbines. Me and my wife's website is the best place to look at them.
http://dairylandwindpower.us

Basically, without wind power here we'd be screwed in the winter. If we didn't have them the generators would have to supply roughly 80% of our power from early December to the end to the February, as our solar can only produce roughly 20% of our power needs in the dead of winter. The wind turbines (three of them at present) sometimes supply 100% of our power during week long windy periods. Other days the generators supply 90%, sometimes for 2-3 days. Our Conext ComBox does not log the wind turbines' input to the system and I don't keep paper logs anymore. But despite that, the turbines supply a pretty big chunk of the power here for three months of the year. I don't even run them in the summer time anymore. They'll be shut down from about the first of June until mid-September or so.

Mark Ross

5th May 2014, 07:55

Using a genset to charge batteries here would be extreme emergency, like the battery down to 70+ DoD and no end in sight to the bad weather. The charger in our inverter is disabled. I can enable it if I ever need it, but I've only ever used it for equalizing.

You can hook up dual generators - one to AC1 and the other to AC2. However, generator support only works on AC2. AC1 has what's called grid support, which is basically the same thing. But there's a caveat with using a generator on AC1 - it has to put out utility grade power. Our Honda EM4000SX has an electronically controlled engine and voltage regulation, with "predictive" electronic voltage and frequency regulation using CT's on the legs and an onboard computer that communicates with the engine ECU. It is a quite expensive generator, it is not an inverter unit so it has exceptional surge capacity, and it does produce utility grade power, even at 125% rated load. It won't work on AC1 with a lesser genset.

So I could basically have a smaller genny (~4kW) on AC2 for generator support then a much larger generator on AC1 if I ever needed to charge the battery bank? What do you mean by "utility grade"? Are we talking the good quality power that comes out of a true sine wave inverter generator?

Thanks! Again we have no decent wind here so just trying to make sure we'd be covered!

EDIT: Also the Honda EM6500 is only like $2700...that doesn't seem real expensive?

Chris Olson

5th May 2014, 08:54

What do you mean by "utility grade"? Are we talking the good quality power that comes out of a true sine wave inverter generator?

EDIT: Also the Honda EM6500 is only like $2700...that doesn't seem real expensive?

Utility grade means the generator has to be able to maintain 60 Hz even when overloaded on surge, +/- 1 Hz. And voltage has to stay at 240 +4/-10. If the generator does not stay in that range at all times on AC1 the inverter will disqualify it, regardless of what you have the acceptable voltage and freq range set to.

The Honda EM-SX generators are quite expensive in the world of gasoline fueled portable gensets. But they are the only ones I've found that can maintain the above specs on output, even operated at 125% of their continuous rated load. Even the residential standby units from Kohler, Briggs, Generac, Cummins, etc., do not maintain the tight tolerances on output that the XW requires on AC1. The Honda EM-SX series does it with precision.

The things have three computers on them - one on the engine that actually controls the engine throttle, fuel mixture and ignition timing, one that processes load information and communicates with the engine ECU, and a third (called iAVR) that handles voltage regulation and also communicates with the load computer and ECU. When a surge load hits the Honda there are CT's on the legs that sense the current requirement. It tells the engine ECU and iAVR about it. The engine ECU responds on the throttle before the freq can drop even .5 Hz, and the iAVR sends the appropriate current to the field before the voltage can "sag" to meet the surge load. The gen head has no exciter winding in it. I have worked on 1.5 MW V-16 Cummins prime generators that cannot the maintain the tolerances on output that those Hondas can. A 3.6 kVA rated genset that can go from 1,500 watts load to 5,000 watts on surge, and hold it for 20 seconds in overload while maintaining perfect 60 Hz and 240V is a pretty amazing feat of engineering. I have no clue why Honda threw that kind of engineering into a portable generator, but they did.

Mark Ross

5th May 2014, 19:25

Can I hook an extended run-time gas tank onto that EM6500 like I do with my EU2000?

I didn't think $2700 was a lot to spend on a generator when I'll be spending north of $6k on solar panels (and probably that much again for racking). But I guess I just saw some Generac at Lowes for like $799 so I guess that is a large difference between brands.

Chris Olson

5th May 2014, 22:13

Can I hook an extended run-time gas tank onto that EM6500 like I do with my EU2000?

You could. But unlike the EU2000, the EM6500 does not have a fuel pump. And it has an evaporative emission control system that would have to be hooked into an external tank. The engine won't run right without the evaporative emission controls hooked up.

I didn't think $2700 was a lot to spend on a generator when I'll be spending north of $6k on solar panels (and probably that much again for racking). But I guess I just saw some Generac at Lowes for like $799 so I guess that is a large difference between brands.

No, absolutely $2,700 isn't a lot to spend. But you're into the price range of smaller low-end diesels. The Generacs are the absolute bottom of the totem pole, bottom feeder, bargain basement, box store el cheapo generators. I mean, they pedal 'em in every home improvement box store that ever hung a shingle so they got market share. But if you want quality, look elsewhere.

Brian McGowan

5th May 2014, 23:59

Chris,
You have probably mentioned it and I went back and looked but I must have missed it so I'll ask, what voltage is your battery bank?
I ask because I am using alternators tied to small engines as a way to keep my batteries up when using the system when the grid goes down and I am not getting enough energy from other sources and let the inverter do all the work. That way I don't have to worry about the generator being accurate enough to match the inverter. When there are heavy surges the battery kicks in to help and when the power is light the alternators supply power enough to run the loads and charge the batteries so they don't get too low.
I am investigating a controller made by a guy that uses an Arduino and running some software he wrote to control the alternator and the engine. One version will actually even start the engine.
The advantages I see are you can run the engine at a slower speed when you are not using much power and run the speed up as needed based on what the temperature of the alternator is and the inverter is what supplies the power so there is only one source of AC and no need to match anything to it.
I am currently using a lawn mower engine because I get them for free on trash night when people put them out in the trash but a while back I acquired a 6hp diesel engine which I am going to use for this so I can burn vegetable oil or biodiesel as well as regular diesel. I want to hook two alternators to it so I can have lots of power if needed.
What are your thoughts on this type of arrangement?

Chris Olson

6th May 2014, 07:44

If you want to use commonly available automotive and marine alternators on a gas or diesel charger, I'd stick with 24V. The 24V alternators (like GM 10SI series) are available for the same price as 12V ones. And if you want a really good alternator that's more efficient that automotive ones, you can get 24V marine 145A alternators for more money. But for 48V you're going to have to buy a pretty expensive ($6,000-7,000) DC genset because nobody makes decent alternators for it in the automotive and marine world.

The only real advantage to 48V is the amount of inverter capacity that's available in one unit without stacking. If you can get by with 4,000 watts of inverter capacity, 24V will work fine.

Rob Beckers

6th May 2014, 09:12

Why has the alternator Voltage any bearing on the battery/inverter Voltage? Doesn't everyone use the chargers that are build into the inverters for doing the charging, i.e. the inverter also takes care of converting 120/240V AC into the requisite battery Voltage?

For anything but the smallest systems there are advantages to using 48V for batteries/inverters: The same (expensive) PV or wind charge controller can handle twice the Wattage, battery wiring can use almost reasonable gauges for high power levels. Once you exceed a handful of PV modules the limitations of even a 24V system (forget about 12V) become clear; the usual 60A charge controller can only handle 2kW (that's just 8 panels these days). With a 24V system you quickly get into 250A breakers and 4/0 wiring. Etc.

Don't get me wrong, I don't want to downplay 12V or 24V systems. They have their place and I sell lots of those. It's just that as soon as you get up to a couple of kW of PV or inverter power the limitations become apparent and I can't tell you how many times I've heard "I wish I made it a 48V system" when customers are looking to expand what they have (and down the line almost everyone expands their system).

-RoB-

Chris Olson

6th May 2014, 09:31

I see nothing wrong with a 24V system up to 4.0 kVA inverter capacity. If you need more inverter capacity than that, then 48V will be better.

The question that I understood was about designing your system to use a homemade DC generator. 48V is pretty hard to find suitable alternators for it. The converted automotive alternators with a permanent magnet rotor don't stand up with the small windings that can fit on their tiny stator core at 48V. Therefore it requires a really expensive purpose-built 48V alternator. You can find them military surplus sometimes, but even those are expensive. If you buy like and Alten Duke DC genset for 48V you'll spend minimum $7 Grand on it. A 24V one can be built with an old lawnmower engine and a $100 10SI alternator. You can even convert a 12V 10SI to 24V by simply changing the rotor in the alternator. But converting to 48V is not doable.

There are some advantages to using DC gensets instead of the inverter's internal charger. We used a DC genset here for over 10 years and they are pretty efficient for combustion engine driven battery charging compared to using an AC generator with an inverter/charger.

Chris Olson

6th May 2014, 12:27

Chris,
I am currently using a lawn mower engine because I get them for free on trash night when people put them out in the trash but a while back I acquired a 6hp diesel engine which I am going to use for this so I can burn vegetable oil or biodiesel as well as regular diesel. I want to hook two alternators to it so I can have lots of power if needed.
What are your thoughts on this type of arrangement?

It took me a bit to find the pictures of this thing. It sat outside the back door of our house going into the power room for several years. It just had a plastic tub upside down over it to keep the snow off it. It had a GM 10SI 44A alternator that I converted to 24V by changing the rotor in it, and I ran it with the 12V regulator full-fielded. It was driven by a 6hp Tecumseh snowblower engine and that engine would start at 30 below zero with no preheat.

https://lh4.googleusercontent.com/-xcqtPUi-A_A/U2kXEJHR86I/AAAAAAAAIHs/2wpLMG8j1u4/w640-h480-no/100_0225.JPG

In the winter time, instead of living like hermits and conserving on every watt we used to start that thing up and leave it idle at about 2200 rpm putting out around 750-800 watts. It would run from supper time to bed time on about 1 quart of gas. We watched TV and movies, made popcorn in the microwave, had all the lights on in the house, even ran Christmas lights with it during Christmas season, and generally lived pretty good thanks to that little gas charger.

With all the money we've spent in the meantime our system is a lot fancier today, and more automatic. But I can't really say that what we had back then was all that inferior to what we have now. It's all a matter of perception I think. I sold that little gas charger for $75 bucks to a young couple that built an off-grid home about 35 miles from us. And they are still using it today, happier than a pig wallowin' in mud because they use it the same way we used it. It allows them to have a more "normal" lifestyle for very little cost. That Tecumseh engine has at least 6,000 hours on it and all I ever did to it was put a new needle and seat in the carb once and change oil in it and replace the spark plug now and then. She may be not be UL Listed, EPA Certified, or even fancy - but I'll tell you what - there's more than one way to skin a cat when it comes to off-grid power.

Brian McGowan

6th May 2014, 23:37

Very nice Chris! I ended up with a vertical shaft engine. Here is the link to the alternator regulator I was talking about.
http://smartdcgenerator.blogspot.com/
I have a 65A alternator and I am controlling it with a big potentiometer which of course requires me to pay attention to it but it keeps me going when the grid goes down and I need power. Although the new batteries made it without the use of any generator. I am still working on getting a new engine on it right now so I will have it if I need it.
I used it a couple of years ago and while the neighbors were going through 5-6 gallons of gas a day I went through 1.5 gallons in 2 days.

Chris Olson

7th May 2014, 08:24

Very nice Chris! I ended up with a vertical shaft engine. Here is the link to the alternator regulator I was talking about.
http://smartdcgenerator.blogspot.com/

It looks like that unit was developed for a marine application, which is kind of cool. We have a sailing yacht but we have solar panels on ours, and no onboard generator.

I used it a couple of years ago and while the neighbors were going through 5-6 gallons of gas a day I went through 1.5 gallons in 2 days.

That's been my experience with DC gensets run as a prime power unit. If you push them hard and try to charge batteries with them, the efficiency suffers as much as using an AC genset with an inverter/charger. But if you use it to manage how fast the battery discharges for long periods of time they are pretty efficient. However, in testing and measuring I've done an AC generator is more efficient for a system with AC loads, and DC is more efficient for a system with DC loads.

It is unfortunate that most people do not understand the concept of prime power for off-grid applications. All they understand is standby power - run the batteries down then start a big genset and recharge them with an inverter/charger. Even though it is horribly inefficient, the concept of using standby power has been established and engrained as "conventional design" by inverter manufacturers putting AC chargers in inverters. However, with a carefully designed prime power system you can cut your fuel burn, powering the same loads, by as much as 43% (measured on our own system that can use either method). Being in the prime and standby generator business for 19 years, I have always been quite interested in this. And "conventional design" using standby power for off-grid is dead wrong. It might be convenient. But convenience never makes up for poor design.

Chris Olson

7th May 2014, 08:40

All they understand is standby power - run the batteries down then start a big genset and recharge them with an inverter/charger. Even though it is horribly inefficient

Hate to quote my own post, but what most people do not understand is that when you use an AC generator to charge batteries with an inverter/charger, only about 65-68% of the power produced by that generator ever makes it to the end loads. The only thing that off-grid folks seem to understand about it is the fuel bill when they have to do it. So the natural tendency is to avoid running the generator at all costs - and "all costs" usually ends up as abusing the batteries, which happen to be the most expensive component in your entire system.

That is the critical flaw in "conventional" design of off-grid power systems. And one of the major reasons why off-grid power is so prohibitively expensive. No matter how much solar panels you throw at the problem, designing a generator into the system is still necessary. How that generator is designed into the system is where it goes to hell in a handbasket.

Mark Ross

7th May 2014, 09:28

Chris, would you say that it's true, that to design a good system with generator support, your total loads have to be pretty well understood? That is where I am having an issue, we will have all sorts of stuff in the house that we have never dealt with before and have unknown power demands, that we will not fully understand until after we have built our place. I wonder if you could suggest any approach I might use to best tackle these unknowns.

I feel like if I get the XW6048 I have the ability to run in your less prefered "conventional" mode or also configured with generator support, so choosing that is a bit of a no brainer. However sizing the battery bank would now appear to be the main issue, to really leverage the benefit of generator support I need to know all my loads so I can minimize expense on the battery bank...?

Appreciate the insights and your incredibly helpful comments! Lots to learn for a newbie like me!

Chris Olson

7th May 2014, 10:21

Knowing what the loads are is imperative. You need to know daily energy consumption in kWh and instead of using convention in sizing the battery bank for 72 hours, cut it to 48 hours. And you need to know what peak loads are to size a peaking generator with the inverter. And you need to know baseline loads during your heavy use periods that don't involve high peaks, but involve continuous discharge at or over the 20 hr rate of the battery bank.

Without that information you won't be able to design it or manage it right.

Any inverter you pick will be able to use the conventional method of charging batteries with a generator. There are only a handful that supply generator support to use a little generator on prime. And there's a breakoff point on your >20hr discharge rate baseline loads where a DC unit will be more efficient on prime than a AC, because the loads aren't high enough to keep a small AC unit loaded at 80% rated load or better.

On that breakoff point, it is possible to use a small inverter genset like a Honda EU2000. The problem with inverter generators is poor surge capacity, and surge capacity is needed on gen support to keep the generator qualified for things like the fridge compressor cycling. So if that baseline is less than about 1600 watts (like during the evenings in the winter when you get home from work and a lots of family activities taking place for 3-4 hours before bedtime), you're on the edge of where a little DC unit just idling and supplying some of the power to reduce the discharge rate of the battery will work better. You're not overloading the inverter during that time, but you're discharging the battery faster than you should be. So you're not going to make it to the next time when you got adequate solar insolation to do a full recharge.

Even on good days in the winter, if the battery has gotten to 50% SOC or below, you may not get adequate solar insolation to get the battery charged after a week to 10 days. So you run the little prime unit for 3-4 hours on a day with bright sunshine to take some of the loads off so the solar can get the job done. Knowing how many kWh to "inject" into the system on a day like that is necessary to make that work. That's what I use our ComBox for. It very accurately logs everything about the system and keeps detailed records for over 10 years. It tells me what to expect from the solar for the day, it tells me what to expect for loads for the day, it shows how many kWh have been put into the battery vs how many kWh has been taken out, and how much we're going to be short. So I run the little prime generator for the exact amount of kWh needed so the RE system can get the battery to 100% SOC.

So you need to live with the system for awhile. Everybody's energy use patterns and habits are different. Some people are home all day and can take advantage of peak solar output. Some work all day and are home only in the evenings so the baseline needs to be managed for 3-4 hours to control battery discharge rate and keep Peukert Effect from reducing how many amp-hours your battery can deliver. I have found you cannot do it with one generator. You need one for peak (inverter overload), and another smaller unit for prime. They both control battery discharge rate. But you can't use the big peaking generator for prime because you can't keep it loaded to its most efficient power output level.

If the system that this thread describes in the first post had a small prime generator that only cost pennies to run the problem with the batteries would have never existed. That tiny generator could have been run on good sunny days in the winter and get the batteries to 100% SOC more often. With only the big generator it is never used until the batteries get too low, then it costs too much to run it to fully recharge the bank. Our own system had the same problem this winter because we were gone for six weeks. We shut down wind turbines, shut off auto-starting the gensets, and just left the system on solar to power the 'fridge, freezer and a 20" box fan we had blowing air to the upstairs from the basement to keep the house above freezing. I don't trust anything that produces power with moving parts left unmonitored for that long. The solar panels got buried under snow, the batteries got deeply discharged and never charged again for at least three weeks. It took awhile to bring them back to 100%. But it shortens the useful life of the battery.

Bryan Couch

7th May 2014, 21:14

So in Mark's case, where he's not sure of all of his loads because the house isn't built yet, he needs to come up with the best possible educated guess at his daily use, peak loads and highest continuous discharge at >20h discharge. Size his bank and generators for these best guesses and build the house and system.

After living with the system for a year or two, maybe three, and monitoring the systems performance and the household use patterns he may (but probably WILL) need to revisit his generator choices and possibly buy a different one or maybe even two and tweak the system for optimum performance.

Am I still following this conversation?

Then.
In using the generator with those heavy continuous load (like the dark winter evening when you're making supper and the kids are doing homework...or watching tv haha) you're saying the little Honda EU2000 will work best if the loads are up over 1600W or so and the larger gen would be there for any large surges (like the fridge)? However if that continuous load was under 1600W a small DC gen would work better because it would be loaded higher than the Honda and therefore run more efficiently?

So really, if one really wanted to make the most of his generator fuel... He would have a large AC generator for the peaks, a small AC generator for the continuous high loads and a second small DC generator for the continuous high loads that are just a little too light to keep the AC unit loaded to around 80%. Or would it make more sense to shut down the little generator at 1500W and wait till the sun comes out the next day and fire it back up to lighten the loads and get a good charge into the batteries?

Sorry if this is repetitious or trivial but I'm new to this and have read enough about your system Chris to know that it is not a typical system but it works extremely well and you have a huge knowledge base that you seem very willing to share. I've learned a lot from your posts.

Bryan

Mark Ross

7th May 2014, 22:00

Yeah I mean at this point I might as well pony up the $2700 or whatever for the Honda 6500, pick up a XW6048, and run some loads off the generator through the inverter as I build to get a sense for how things are shaping up...? Once I have more info I could then buy the battery bank, panels, etc best sized to match the system? I'm still a little unclear as to a best approach to move forward.

Chris Olson

7th May 2014, 22:20

Bryan, that's it in sort of a nutshell. There's some considerations;
1. Batteries are typically rated on amp-hour capacity (and therefore kWh capacity) at the 20 hr rate of discharge.

2. If the battery discharges at the 20 hr rate you only got 10 hours from fully charged to 50% SOC.

3. If you discharge the battery slower than the 20 hr rate the Peukert Effect does not gain you substantial amp-hour capacity.

4. If you discharge faster than the 20 hr rate the Peukert Effect usually reduces your amp-hours available to loads by a factor of 1.25 - 1.26.

Discharging your battery faster than the 20 hour rate must be avoided on long cycles where you have insufficient solar, or you'll end up charging batteries with a generator. To make it two days on battery power alone, from 100% SOC to 50% SOC, your average discharge cannot exceed the 100 hr rate.

OK, so let's put some figures to this. If you have a 1,000 amp-hour battery bank @ 48V nominal, the 100 hr discharge rate is 10 amps. That's 480 watts @ 48V nominal and about 460 watts to your loads. At 460 watts average load you'll be using only 11 kWh/day in your off-grid home.

So, this 11 kWh/day off-grid home runs into a period of bad weather in the middle of winter where the solar panels only produce a 2-3 kWh/day for four days, in snowy overcast conditions. What are your choices? Not much. You're going to be using a generator (unless you got wind turbines or a hydro setup).

How you use the generator determines how much it will cost you. Internal combustion engines have to run near the peak of their torque curve for best BSFC (Brake Specific Fuel Consumption). With an AC generator that means 80% rated load, or better. Generators that can be throttled (AC inverter units, and DC units) are different - the peak torque changes from WOT to partial throttle conditions so you can achieve higher efficiency at lower load (but not as high as at WOT and peak torque).

To get every last possible kWh to loads from your generator fuel you need to manage this loading. Managing it with a battery charger is not efficient because you have losses in the battery charger (10%), you have losses in battery charging (15%), and you have losses in converting the stored energy in the battery back to AC for the loads (~8%). It should be obvious that you can skip all these losses in battery charging by using the generator's power direct to loads instead. It does the same thing as battery charging - injects power into the system from the generator. But it does it more efficiently.

The smallest practical AC generator is about 1,600 watts with a 1,200 watt prime load on it. Anything lesser you will be relegated to a DC unit for prime power, and it will not be as efficient on a system with AC loads as an AC generator will be. But it will be more efficient than using your AC generator to charge batteries.

Just a warning on these little inverter generators - our off-grid neighbors that live 4 miles thru the trees from us have a XW6048 and a Perkins diesel for standby power. They bought a little Honda EU2000 and a Outback transformer for prime power. The little Honda cannot put out enough surge power to magnetize the core in that big 6.0 kVA transformer. The inrush to the transformer is 58 amps when you throw the breaker and it causes the inverter on the little genset to fault and shut down. Our conventional Honda generator (EP2500CX - 2.0 kVA prime rating) handles it with no problem.

I built the neighbor a soft start for their transformer out of some water heater elements and a bypass switch. The water heater elements are inline resistors that reduce the voltage to the transformer when the core magnetizes so it reduces the inrush to where the little EU2000 can handle it. Once the core is magnetized then it works fine and the resistors can be bypassed.

On the final part about sizing things - it's no different than buying a truck. Do you buy a semi to haul a 200 lb load of rocks? No. A pickup truck will work fine. You have to know what you plan on hauling with it before you buy it.

An off-grid power system is the same thing - you can't even begin to know what inverter to buy, or how much batteries, or how much generator, until you know what your loads are. If you don't know that, or can't figure it out, I would highly suggest buying a pickup rather than a semi, then living with it for awhile until you get a good handle on what truck you really need.

Brian McGowan

7th May 2014, 23:19

[QUOTE=Chris Olson;26506]It looks like that unit was developed for a marine application, which is kind of cool. We have a sailing yacht but we have solar panels on ours, and no onboard generator.

Sailing vessel or not it has the correct charging profile for off grid systems as well.
I worked with a guy that developed an inverter for rental sailboats in the Caribbean. It was designed to kick over and power a 1/2HP refrigeration unit. One of those "blue ice" units. We used to profile and match the power transistors so they would take as equal parts of the current as was possible. The feed wires to the transistors were cut to the exact same length for the same reason. The other feature I always thought was cool about the unit was the fact that in the "standby" mode it would put 12 volts on the AC circuit it was attached to and literally stay off until a switch was closed on an appliance at which point it would fire up and dive the load until the load current dropped off and it would go back to the "standby" mode. It used microamps while it was in standby. I am pretty sure I still have the board layout and part list from that unit.

[That's been my experience with DC gensets run as a prime power unit..........]
When I was doing this to keep myself alive during power outages I was using solar during the day and the generator for a few hours at night. During the day the panels would generally supply more than I needed thus running the appliances and charging the batteries and at night I would set the generator up to provide about 15 amps more than the average load. When the load increased the batteries would help supply the extra current and when the heavy load subsided the generator would go back to charging the batteries as well as handling the lesser load. Basically the batteries were acting like a really big capacitor. Never getting very discharged but helping to smooth out the load. They never had to supply the whole load so I didn't exceed their discharge rate. The entertainment center, computer room and some lights were the steady load and the fridge kicking on and off was the increased/variable load.

Chris Olson

8th May 2014, 07:28

[QUOTE=Chris Olson;26506]Basically the batteries were acting like a really big capacitor. Never getting very discharged but helping to smooth out the load. They never had to supply the whole load so I didn't exceed their discharge rate. The entertainment center, computer room and some lights were the steady load and the fridge kicking on and off was the increased/variable load.

That's a good way to describe it. We used our DC generator for years to do the same thing. On a system with AC loads our efficiency is a little better with an AC generator because we're skipping the losses in the inverter in powering the loads with the generator. The caveat is that the AC generator requires a minimum load in order for it to work. The DC unit is more infinitely adjustable, so it will handle a wider range of loading at pretty high efficiency.

With our system, where we use 20-25 kWh/day in the winter, the little AC generator works good because our high draw times in the evenings run at 2,000-3,000 watts for 4-5 hours. So the system can be set to pull 2.0 kVA from the generator and the battery supplies the rest, which keeps the discharge rate of the battery < 20hr rate and simultaneously loads the little generator to a fuel burn of .17gal/hr. At that fuel burn, it costs us 62 cents/hr to run the little generator (fuel at $3.64/gallon) for 1.89 kWh (with Power Factor). So our prime power costs us 33 cents/kWh out-of-pocket.

The best I have ever been able to achieve using a generator for battery charging with an inverter/charger is 68 cents/kWh to the end loads. And that's if I stop the generator when the batteries reach 80% SOC and don't gas the battery with it. If you gas the battery (absorb it) with a genset and inverter/charger the cost to the end loads is so ridiculous that I don't see how anybody could afford it long term.

The downside to all this is that it requires management of the system to decide when to run the generator, and how much. It's easy to set that prime generator up in the evenings, start it and let the system do it's thing - but it's also easy to forget that it's running. I tend to suddenly remember, hey that little generator is running yet at 10:30PM. I look at the system status and our high draw loads have subsided to 1.4 kVA and the little generator is carrying all of it and the batteries are idle and no discharge at all. And the generator is no longer loaded to its most efficient power level. I have a friend that lives off-grid in the Yukon that understands all this and has developed a battery management system that starts and stops his prime generator with a XW inverter and AGS using one of the thermostat inputs on the AGS to signal the genset. But he doesn't have it perfected yet. The only fully automated system that I know of that does this (for DC loads) is Sandia's ACONF unit.