No announcement yet.

Sizing Off-Grid Systems And Using Gen Support

This is a sticky topic.
  • Filter
  • Time
  • Show
Clear All
new posts

  • Sizing Off-Grid Systems And Using Gen Support

    I'd like to try to make a contribution on what seems to be a very hard thing to understand for most folks looking at making a move to off-grid. This is more applicable to medium to larger full-time off-grid homes. Smaller systems like single 120V Outback setups in part-time off-grid cabins, or very small off-grid full time homes where daily energy consumption is under 5 kWh are probably not going to benefit from it.

    I am dealing with sizing of the inverter, batteries and generator(s). Sizing battery banks and solar/wind/hydro capacity is a totally different issue. But the basis of my contribution is to save money on batteries and inverters and generator fuel by properly sizing the generator and inverter to meet loads, and using Generator Support to prevent having to buy expensive stacked inverter setups and the additional battery capacity required to power them at full rated load. Stacked inverters might look impressive on the wall. But if you don't have the batteries to power them they're useless. And few off-grid folks can afford the batteries it takes to power even ONE 6.0 kVA inverter at full rated load for more than about one hour.

    There's two schools of thought on off-grid. The one assumes off-grid means never running a generator unless in extreme emergency, and using a generator is a "sin". The other (more targeted to larger systems) integrates the generator with the system with auto-start, etc. and it is run on a regular basis to enjoy a more comfortable off-grid lifestyle. If you think making the move to off-grid is "going green" you are dreaming. You only make the move to off-grid because you can't get utility power where you want to live. It has nothing to do with "going green" except for the fact that it takes a lot of green (money) to do it.

    Common mistakes I see made all the time:
    1.) Buying a generator that is way too big. The result is that you can't afford to run it because it sucks OPEC dry as soon as you turn the key. A combustion engine generator needs to be operated at 80% rated load or better for least cost/kWh. If you buy a 8.0 kVA generator do you have better than 6.0 kVA load for it when it is running? Few do - even with larger systems.

    2.) Installing stacked inverters for intermittent heavy loads when a single one with generator support will do the job.

    3.) Purchasing excessive battery storage capacity (and not near enough RE capacity to charge them, which is a separate topic), thinking you should be able to go three days without running a generator if there's no sun. This is ridiculous. 24 hours is adequate on straight battery power if you design a system using a tiny prime generator to manage how much you use from batteries during peak draw times (like in the evenings in the winter). The result is that you only need 1/3 of the battery capacity you think you need by managing loads with a little genset to keep the peak draw on the batteries below the 20 hr rate of discharge. Put the pencil to it over the expected life of batteries and you will find that even with fuel costing twice what it is presently in the US, the fuel in a tiny generator to manage battery capacity is cheaper per kWh than the battery replacement cost at 7 years.

    Off-grid power is VERY expensive. Much more expensive than utility power. Batteries are the single most expensive component of an off-grid power system. Not inverters, generators, fuel, solar panels or wind turbines. Batteries. I have seen more "hobby" hybrid grid/off-grid systems get abandoned when the batteries failed than for any other reason. Having to spend $3,000+ on a battery bank at replacement time becomes a reminder of the realities of how much off-grid power really costs.

    There is a common misconception that you cannot cycle batteries below 50% DoD or it will ruin them. This is totally and blatantly false for real deep-cycle batteries. They can regularly be discharged to 80% DoD and it doesn't hurt them one bit. And in fact, it makes them more efficient on charging.

    The rules of thumb I have developed over 14 years of doing this:

    1.) Purchase the battery capacity you need to power your average daily consumption with the batteries discharged to 80% DoD over 24 hours. Example: you use 15 kWh/day in your home on average. You need a 19 kWh battery bank. Normal cycle depth will be around 50% on the average day if you have adequate RE generating capacity to recharge them daily (under average solar insolation conditions) plus carry normal loads during the recharge period. On poor RE days you will cycle them deeper and manage how much you use from the battery by using your little prime genset. Minimizing the size of the battery bank up front prevents "sticker shock" when the day comes that you have to replace them.

    2.) Determine what your peak load will be and buy an inverter that can carry that peak load for a few minutes, then switch to using Gen Support to power the peak load.

    3.) Size your peaking generator so it reduces the load on the inverter during peak load to no more than the 10hr discharge rate of the battery bank. Example: your peak load is 7.5 kVA. You have a 6.0 kVA inverter that can handle the 7.5 kVA peak for 15 minutes before the load on it must be reduced to below its rated continuous output. You have a 1,000ah battery bank at the 20hr rate (40 amp draw on the bank to 80% DoD). The 10hr rate will be approximately 70 amp draw on the bank with Peukert Effect. 70A draw on a 48V system @ 90% inverter efficiency is 3.0 kVA. You need a generator capable of delivering 4.5-5.0 kVA.

    And this brings up the topic of Generator Support, which is the only way all the above works. I have talked to many people who think their Magnum PAEs et al have it. They don't. With an inverter that does NOT have generator support in it, you are relegated to purchasing an excessively over-sized generator to be able to carry the peak loads because the inverter does not have the capability to sync with the genset and assist it. This does not give you the ability to carry a wide variety of loads and always keep the genset at peak efficiency (read above about sucking OPEC dry). Nor can you even come CLOSE to efficiently use this big genset for battery charging if you charge batteries with it thru the inverter/charger. There's only a handful of off-grid inverters that have Generator Support. And these inverters are quite expensive compared to the ones that don't have it. Carefully research what you're buying for an inverter before you buy it. While the batteries are the most expensive thing, the inverter is the heart of it all that makes it work.

    So how does peak load support using this system work, exactly, and what does it do for you? In summary:
    1.) It conserves battery energy during peak load events on your system by reducing the amp draw on your battery so the Peukert Effect doesn't sack them out.
    2.) It allows the use of a smaller more efficient generator that is able to be kept at >80% rated load at all times
    3.) It saves money on batteries and inverters by not having to buy twice as much for stacked inverter setups.

    I made a demonstration video of it with our XW Power System to show you what it does in a real life off-grid home where it is used every single day. And has been used here for many years, previously with our Xantrex SW Plus before we replaced it with the XW system:

    off-grid in Northern Wisconsin for 14 years

  • #2
    Interesting read. Need to think on it so as to digest it well. I did not realize there were off-grid inverters that actually synched up with a generator. I have a magnum PAE that you reference and essentially the batteries run to a predetermined point then the generator starts. The inverter goes off-line while the generator gives pass through power while charging the batteries back up.

    So the main advantage of the synch allows you to increase peak load? Smaller inverter / smaller battery bank / smaller generator? If so, is it the load AND battery status that starts the generator? Do you have a seperate charger for the bank?


    • #3
      Thanks for this great introduction. Someday, if and when I get around to building my cabin in the woods, this will be very helpful to me.

      One topic that probably should also be addressed is demand. Since storage is the most expensive component of an off-grid system, it makes sense to move as much demand as possible to the times when the sun is shining.

      Other than the obvious (run the dishwasher during the day, etc.), do you have any strategies or tools for managing demand?
      16x TenK 410W modules + 14x TenK 500W inverters


      • #4
        Robert1234 -
        The XW-AGS has four load start triggers. One of them is based on the sum of the current on L1 and L2 and is a VA measurement to determine actual load on the system. I always use VA or kVA instead of watts because to convert from watts to VA on AC requires use of the Power Factor. Watts is amps x volts on DC. VA is amps x volts on AC.

        The VA setting (or Load Start Amps) prevents long term inverter overload.

        The other three are based on battery voltage and are applied for long term high continuous loads that do not exceed the VA setting. A timer is applied on the three battery voltage based load start triggers. There is a 2hr timer for intermediate long term loads, a 15min timer for higher long term loads, and a 30 second timer for loads that almost approach overload but not quite. All three of the voltage-based triggers are cleared based on battery bank voltage recovery to a level that you set that stops the generator without charging the batteries when the load goes off.

        We do not charge batteries with gensets. They are only used for load management. The charger in our XW is set to Disabled so the gensets cannot use it. The only thing that charges our batteries is either solar or wind power. We only resort to using a genset for battery charging in extreme emergency like when we recently came home from sailing in the Caribbean for a couple months and our system is pretty much dead to the world with the solar panels buried under 6 feet of snow and wind turbines froze up. We need to get it back online. So I enabled the charger in the XW, got one generator started, and used it to bring the batteries up and carry loads until I got the rest of the system working again.

        pleppik -
        Scheduling loads is one way to use peak solar power during the day. However, it is not always possible. We are many times out working in the field during the day, or gone doing other things. We designed our system on purpose to be able to do what we want when we want to do it. And we are not afraid to have to use a genset to make it happen. It is about convenience.

        We get home and my wife wants to do some laundry and run her electric clothes dryer, or we want to make a pizza in the oven for supper. Or I want to do some welding in my shop with my Lincoln 225A spark welder at 9:00 at night. So we just do it and if it causes the peaking generator to start and run for a half hour or whatever, so be it. The system was designed to be able to handle the load at any time totally automatic. There are some things we can't do like run the electric clothes dryer and oven at the same time. Or I can't do any welding when either the oven or electric clothes dryer are operating. But we work around that because they are all short period very high draw loads. But that is the only load scheduling we do.
        off-grid in Northern Wisconsin for 14 years


        • #5
          Originally posted by ChrisOlson View Post
          Scheduling loads is one way to use peak solar power during the day. However, it is not always possible. We are many times out working in the field during the day, or gone doing other things. We designed our system on purpose to be able to do what we want when we want to do it. And we are not afraid to have to use a genset to make it happen. It is about convenience.
          Have you explored trying to recover waste heat from the generator to supplement your home heating? I know there are some commercial "Combined Heat and Power" units on the market--not very common in the U.S., but apparently more popular in some European countries. It seems that it shouldn't be too hard to build a simple system to collect at least some of the heat from the generator.

          I would guess that if you use propane for heat in the winter, you probably spend more on that than you spend on either batteries or fuel for the generator.
          16x TenK 410W modules + 14x TenK 500W inverters


          • #6
            Originally posted by pleppik View Post
            Have you explored trying to recover waste heat from the generator to supplement your home heating?
            No. We heat 100% with wood in a central forced air Daka furnace (made in Pine City, MN) that has a catalytic recombustor on it to burn the particulates in the wood smoke. We have a virtually endless wood supply here that will last us for over 1,500 years just doing select cutting of mature hardwoods on our own property. There is no way we can keep up cutting the mature trees for firewood to open up sunlight area for the new undergrowth before they die. I have probably 3 years worth of harvested logs that I haven't processed, but got out of the woods because if I don't they will start to rot and go bad. I sell processed firewood and chips to about 30 customers in the area and still can't keep up with the rate that hardwoods reach maturity and are ready for harvest on our land.

            It is not worth it to mess with CHP when we have a fuel source like that.

            We do not burn any propane at all, nor will my wife allow a propane line into the house. Years ago we had an explosion in a rented house we lived in that blew one wall and all the windows out and lifted the house off its foundation and shifted it about 2 feet. It was due to a valve on the water heater that malfunctioned and we lost most of our personal belongings from it. We had a propane genset for a couple years and my wife made me hide the tank for that behind a rock wall and bunch of pine trees where she couldn't see it. But it bothered her to no end that it was there. And the propane genset didn't work worth a crap anyway in cold weather so we got rid of it.

            IMO, propane is one of the worst choices for off-grid fuel. You couldn't even buy the stuff around here this winter because all the suppliers ran out. And the last they had when they were rationing to people for fillups, they were selling for over $7/gallon. Plus there is no way in hell a LP truck can get in here. Our 500 gallon tank for the genset used to be mounted on a wagon running gear and I towed it 25 miles to where the truck could come and fill it up, then towed it home. The only upside to propane is its shelf-life. But it has so many downsides that it's not even worth considering unless you live in a tropical climate like Missouri. I can haul a 500 gallon tank of diesel fuel in here and it's got one hell of a lot more energy in that 500 gallons of diesel fuel than 500 gallons of propane has, for less than half the cost per BTU.

            The only thing we use the waste heat from the gensets for is to heat the powerhouse in the winter. The powerhouse has thermostatically controlled ventilation that keeps it at 75F in there in the winter for the air-cooled engines so they don't run too cold. Never did get into using liquid-cooled genset because it takes too damn much time to warm them up properly before they can accept full rated load.
            off-grid in Northern Wisconsin for 14 years


            • #7
              Originally posted by ChrisOlson View Post
              There is a common misconception that you cannot cycle batteries below 50% DoD or it will ruin them. This is totally and blatantly false for real deep-cycle batteries. They can regularly be discharged to 80% DoD and it doesn't hurt them one bit. And in fact, it makes them more efficient on charging.
              Sorry got to bust you on this one as being absolutely false. There is not one bit of credible data to support that statement and contrary to every bit of published data by the manufactures and third party testing. If folks go discharging to 80% DOD are asking for very short battery life, It will take that 5 year battery and turn it into 2 years or less. Just about every DOD vs Cycle life graph published looks like this.

              MSEE, PE


              • #8
                Originally posted by Sunking View Post
                If folks go discharging to 80% DOD are asking for very short battery life, It will take that 5 year battery and turn it into 2 years or less. Just about every DOD vs Cycle life graph published looks like this.
                Been doing it to ours for five years come March 21 and not a single problem with it. I have very sophisticated equipment on our system that keeps track of every kWh into and out of the batteries and they still have very close to rated capacity on discharge.

                It's all about cost/kWh. You can cycle a battery shallow and get more more years from it. But that doesn't mean you got more kWh into and out of that battery. Or you can buy less batteries and work the snot out of them and replace them more often, getting better charging efficiency and more kWh into and out of them over their life. I prefer the latter method. Getting 7-8 years down the road and dealing with batteries that only have 50% original capacity left and trying to squeeze another couple years out of them when they're half dead doesn't save you any money because their charging efficiency at that age is going up in smoke. You're throwing your energy you put into them right out the window.

                We also use quite sophisticated methods to reduce cycling. We don't recharge every day and waste a cycle in the battery's life. During periods of poor RE production our batteries will sometimes go thru a gradual 7-10 day decline to a very low SOC before being recharged again. That is only one cycle over that 7-10 day period.

                Not every manufacturer recommends cycling below 50% SOC. OTOH, some specifically recommend it, like GB Industrial, who doesn't recommend recharging their forklift batteries at ALL until they reach 80% DoD. Rolls specifically recommends deeply discharging shallow cycled batteries at least once per month to keep them healthy.

                As usual, check with your battery manufacturer.

                2. Over Charging and Opportunity Charging
                Industrial batteries are typically designed to last at least 1,500 charge cycles, over a five to fifteen year period. Each time you charge a battery, regardless of how long, it constitutes one cycle.

                Consistently charging a battery twice per day, during lunch breaks for example, is known as Opportunity Charging, and reduces the useful life of a battery by 50%.

                The additional heat generated by opportunity charging a battery usually reduces the run time equal or greater in proportion to the amount of charge it actually received, making the practice completely ineffective and costly.

                Routinely charging the battery before it is 80% discharged is another common form of over charging. For example, if you only use the battery a few hours a day, it’s best to use it until it is truly in need of charging before actually plugging it in. Remember, each charge constitutes one cycle, so try not to charge unnecessarily.


                And here is Rolls' cycle depth vs cycles for the 5000-series. At 80% DoD you are over 5 years life assuming you would cycle that battery every single day. Our batteries, according to our logging data, get cycled about 112 times per year with our present management regime. At their published number of cycles to 80% DoD we should expect 17 years from them. In reality our goal is between 7-10:
                off-grid in Northern Wisconsin for 14 years


                • #9
                  Originally posted by Sunking View Post

                  Thanks for providing a "typical" battery cycle life graph that we can look at.

                  The first thing that you see is that at 20% DOD, the number of cycles is about ~2500.
                  At 80% DOD, the number of cycles drops drastically to only ~500.
                  The first reaction to that is "Oh my, the life in cycles is only one fifth of what it would be with the lower discharge level."
                  And that would be perfectly correct if what you are looking at is a system that is designed to draw the batteries down to 80% DOD every day.
                  There are two other ways to look at the lesson that this graph is teaching us:

                  1. If you plan for 20% DOD and accept four days of no sun before being forced to start the generator or conserve drastically, then you are at 80% DOD and find that you have used up the battery life equivalent to five single day 20% cycles. But that battery life increment has carried you through four days without running the generator.
                  No longer the simple factor of five loss of battery life that we took from the first look.

                  2. If instead of designing for 20% DOD and getting ~2500 daily cycles, we design for 80% daily DOD, we will be replacing out batteries five times as often. But we are only buying a battery bank that is one fourth the size. Still not a good deal and probably a bad way to design if you want reserve capacity and autonomy, but not the factor of five misjudgement that we thought we saw at first glance.

                  So, if we assume that we are just talking about batteries which follow that sample curve exactly, we find that higher discharge depth does cost system life, but is still withing the range of an engineering tradeoff rather than a factor of five mistake.

                  What I have not seen so far is a graph of the effect of say, 10% of the time using 80%DOD cycles mixed with 90% of the time using 20%DOD cycles. That mix sounds more like what Chris is discussing in his last point. He is still recommending as design DOD of only 20%.
                  SunnyBoy 3000 US, 18 BP Solar 175B panels.


                  • #10
                    1) I'm making this a stickie. [done]

                    2) because it's a stickie, and of issues we've had before, all replies are going to be closely monitored, and the "me too's" will be purged to allow focus to the meat of the issue.
                    Powerfab top of pole PV mount (2) | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
                    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
                    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A



                    • #11
                      Well a 5 day reserve only works out to 2.5 days of autonomy so as not to go below 50% DOD. Once below 50% lead sulfate crystals begin to harden and you will not be able to dissolve all of them with a full charge. The deeper you discharge, and the longer you leave them discharge accelerates the process. 95% of all battery failures are sulfated batteries. Every well cared for battery will die from sulfation. The question is how long before that happens and how to delay it.

                      Chris don't get your shorts in a knot, I agree with a lot of what you are saying, just not on the batteries, and might have a few minor issues on generator sizing.
                      MSEE, PE


                      • #12
                        Originally posted by Sunking View Post
                        Chris don't get your shorts in a knot, I agree with a lot of what you are saying, just not on the batteries, and might have a few minor issues on generator sizing.
                        Sunking - no worries there. I don't even wear shorts

                        I should qualify and quantify what I am saying. If you buy Trojan T-105's and cycle them to 80% DoD they're going to be shot in 2 years. Of that there is little doubt. But we do not have light duty batteries here. We bought The Best There Is when we bought batteries. And that is why I said in my original post "real deep cycles". I have seen time after time where somebody says if cycle your batteries below 50% SOC they will be ruined for life. This is not true. Maybe for some real light duty hybrid deep cycles you buy off the shelf at Walmart. But not for Rolls 5000's.

                        I went thru all the math on this when I originally figured it out (I had a spreadsheet that I designed to figure it out and if I can find that I'll post it), and how many kWh you get into and out of the batteries vs cost when it comes time to replace them. And that's why we manage our system the way we do. It has been working well here because the health certificate is issued when you have to discharge them to 80% DoD and see what they actually got in them. And at 5 years ours pass.

                        As inetdog mentioned, each of our cycles does not go to 80% DoD either. Some are shallow cycles and I hate those because they seem to be hard on the battery. If we go for a couple weeks in the summer and the batteries get shallow cycled like to 20-30% DoD they lose capacity on the first discharge to 80% DoD. We don't get that capacity back until we pull 'em almost dead and then fully recharge. My logging data shows it plain as day. These big tall suckers need to be boiled to keep them mixed up or the dense electrolyte starts to settle to the bottom and they go "flat". Only way to do that is pull them down until the charge cycle can actually do something about keeping the battery mixed up and healthy.

                        So I see it as a multifaceted issue that has a lot more to do with it than how many (theoretical) years you can get by shallow cycling vs deep cycling. When I figured it out I went with cost/kWh over the expected life.
                        off-grid in Northern Wisconsin for 14 years


                        • #13
                          I'm getting questions from folks who have seen the video on auto-starting gensets and what inverters to use. This is a completely different topic but since it does involve integration with the system for automatic handling of peak load events on off-grid systems, I'll provide a brief rundown of what I know about.

                          Both Schneider and Magnum Energy build programmable AGS that integrates with the system and will start a genset based on system loading. These controllers can can be used with three-wire or two-wire generators. Of the two, Schneider has 120/240V split-phase inverters (both XW and Conext SW series), Magnum Energy has a MSH4024RE 120V single-phase inverter available with generator support. They are both fairly simple to set up with three-wire gensets, or two-wire, and they are both proven to work.

                          In older inverters, the Xantrex SW/SW Plus both have generator support, and are proven over many years to work flawlessly with all types of gensets using either the auto-start relays inside the SW, or the external GSM (Generator Start Module) used with the SW Plus.

                          Outback does not build a programmable AGS that integrates with their systems. With an Outback all you have to auto-start a genset is a dry contact Aux port. Since the vast majority of generators used on off-grid power are three-wire, with an Outback you have to buy a pretty expensive Atkinson generator controller to convert three-wire generators to two-wire start.

                          The Outback GS8048 Radian is advertised with generator support. However the Radian is primarily designed as a grid-tie inverter. I spent most of one afternoon trying to get one to work off-grid with a 4.0 kVA Perkins diesel genset and it did not work. It would qualify the genset, attempt to load it, then immediately spit it off and repeat. We tried setting allowable voltage and freq range down as low as they would go, to no avail. We were told the genset is the problem so we loaded it up in the back of my Dodge Cummins and hauled it to my place, hooked it up our system to try it on the XW. It worked perfectly fine on the XW. We never did get it to work on the Radian.

                          We opened the Radian up and looked at it and they forgot one critical hardware component for gen support. Instead of a center-tapped transformer to assist in balancing genset legs during initial loading they have two paralleled split-phase modules that are basically FX units. If you recall, the FX-series was supposed to have gen support way back when, but it never got done. Gen support on 120V single-phase is pretty easy to do. On split-phase it isn't. Xantrex has years of experience building split-phase inverters. The Radian is Outback's first attempt at one. So it may work on the Radian with an excessively over-sized genset or with a PSX-240 transformer on the AC2 input. We did not have either handy for testing on that particular day. And the owner got disgusted with it and sold it and bought a XW6048. So I am not able to work with it anymore to see what it takes. I was hoping to test it by taking our PSX-240 to the neighbor's place but there was a lot of swearing, screwdrivers and pliers flying the air and wires being ripped out to put in a different inverter before I got there, so it never happened.

                          The Outback GVFX-series have Grid Support and some folks have fiddled with them and gotten a rudimentary form of Gen Support to work using inverter type gensets. However, Outback does not even recommend using a generator on a GVFX at all, and they are not recommended for off-grid use. They are grid-tie inverters.

                          Supposedly, SMA inverters have generator support in them. They are not as common in the US as Schneider, Magnum or Outback and I have never tested one, or heard of anybody using one with generator support. If somebody owns own and has tested it, it would be nice to hear about it.

                          You can use a manually started genset for peak load off-grid power. But it requires user intervention in the form of first manually starting the genset and bringing it online before turning on your high-draw load. And it requires remembering to shut the genset off when your high-draw load is done.

                          So at the present, the list of off-grid inverters that have gen support is pretty small:
                          In 120V single-phase:
                          Xantrex SW or SW Plus (proven over 20+ years)
                          Magnum Energy MSH4024RE (fairly new but I have two reports of them being used successfully with generator support)
                          SMA Sunny Island (is supposed to have it but I have never heard of anybody using it or testing it)

                          In 120/240V split-phase:
                          Schneider Conext XW-series (proven over many years)
                          Schneider Conext SW-series (fairly new and I have one report of being used successfully with generator support off-grid with a CPE generator)
                          Outback GS8048 Radian (fairly new and may require a balancing transformer, and may work with an excessively over-sized genset that can handle high split-phase leg imbalances).
                          off-grid in Northern Wisconsin for 14 years


                          • #14
                            Originally posted by Sunking View Post
                            Well a 5 day reserve only works out to 2.5 days of autonomy so as not to go below 50% DOD. Once below 50% lead sulfate crystals begin to harden and you will not be able to dissolve all of them with a full charge.
                            I want to add some information to this thread to show folks how this all works in a real-world installation that does not use theories and instead uses logged data to prove it. They say pictures are worth a thousand words.

                            We started the day's logging at 12:00AM with our battery at 88% SOC. We went all night using very little power and woke up this morning to snow and freezing rain coming down. The weather forecast is not good for the next 4 days at least. We never seen the sun all day and never got much from wind either.

                            What do most off-grid people do when this happens? They continue on battery power until the point comes where they have to use a genset and inverter/charger to charge the battery back up. We don't. We went until noon and then decided to start and run our little prime genset all afternoon at 1.8 kVA generator support setting. The little generator will carry loads up to 1.8 kVA, then the inverter starts helping it from battery power if the load goes above that. When the load drops below 1.8 kVA, no energy is used from the battery and the load is carried by the generator. The generator ran for 4 hours, 58 minutes and burned .87 gallons of gas, which cost us $3.08

                            So what did it achieve? This is the graphed data showing loads vs battery discharge for the day up to the 7:00 PM hour. If the graph is blown up the battery discharge line is very slightly above the load line thru the night time hours, even though they look the same on this small graph pulled from the tablet computer:

                            Just before lunch time we made the decision to start the little genset and run it for about 7 kWh today. This is the loads vs genset input to the system. The kVA (actual load) setting for the generator is always higher than kW shown on the graph because of lagging Power Factor.

                            This is the prime genset output curve for the day, and gen stats showing we used 6.9 kWh from it for just about 5 hours. Our cost for this power today was about 44 cents/kWh as we had abnormally low loads all day. Now, 44 cents/kWh is expensive by utility power standards. But it is pretty cheap by off-grid standards:

                            The net result:
                            Our battery has dropped from 88% SOC at midnight to 79% SOC tonight as I write this. If we wouldn't have used the little prime genset today, instead of the battery being at 79% SOC like it is right now, it would be at only 64% SOC. Tomorrow would be another poor day and sometime during the day we would be running a genset to charge batteries. And I got data that shows you how much that costs per kWh - and we don't even want to go there.

                            Are you starting to see how it all works? Instead of cycling our battery down and recharging with a genset, we use that little genset to manage how fast the battery discharges during poor conditions. We can go 7-10 days doing this without ever using a genset to charge batteries at a cost of 35 - 45 cents/kWh, depending on how heavily we can load that little generator. Today, due to the lighter loads, it is a poor example of cost effectiveness because the prime power is more expensive when the genset is lightly loaded. But it is still a good example of how it all works.

                            We can do this for a week to 10 days, starting with a full battery, and only run one discharge/charge cycle on those batteries for the whole period of poor weather. Whether you agree with it or not, it works. Been doing it for years - many years with a DC genset and fewer with AC genset. But it still works. Many folks can't get their head wrapped around the concept because they've been taught that "the way you do it" is to charge batteries with a genset and cycle them several times, instead of managing their discharge rate with the genset and only cycling them once.

                            So I hope this helps to visualize how it works, or gives a better understanding of it. It goes against all convention in design of off-grid power systems. But just because all the books say "this is how you do it" doesn't mean the books cover all ways to do it. As an engineer I spent a good portion of my career being expected to think outside the box.
                            Last edited by ChrisOlson; 03-17-2014, 11:45 PM. Reason: spelling, extra attachment
                            off-grid in Northern Wisconsin for 14 years


                            • #15
                              Following up on the DOD discussion.... (Mike if after this is answered you feel it should be deleted from the sticky as being off-topic, by all means do so but this has been running around in my head and it's as good a time as any to ask about it.)

                              If I take the typical battery curve & multiply the DOD & the number of expected cycles, it seems that the total kwh available over the life of the battery bank divided by the total investment is pretty close to a constant. Is this true? (I realize my "typical curve" may be not be totally accurate, but it looks fairly close to what SunKing displayed earlier.)

                              My point in asking is that in a higher DOD setup the initial fixed investment is lower and the total (long term) investment with regards to battery consumption may be essentially the same as that of a low DOD design. (Battery replacement less often at 20% DOD, but it will also be at a much higher cost due to the increased bank sizing).