Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Building an "off grid" facility ran off large 3 phase genny. Need batteries and inverter thoughts.

Not open for further replies.


Getting ready to throw a building up to handle some industrial work we have going. It was not up for consideration previous because we could not get power to it economically. At this point some numbers have been run that make a large diesel generator the best option. The power from it won't be as cheap or easy as grid power but we have to work with it.

Basically this needs to be a generator/inverter hybrid system in which while the generator is running during the day to run machines and do work, it needs to charge a large battery bank. The battery system is primarily so the building still has "some power" and if the generator fails, there is a little redundancy for small loads like OH lighting. I will also have an air conditioner in the office areas which is a small load but will certainly pull on the battery system. It needs to be able to run all night on the battery bank.

In the winter, the plan is to recover all the waste heat from the generator into a water:water heat exchange tank where the heat will be directed into the radiant floor heat tubing system. We will leave the radiator on the generator and install a Tstat controlled electric fan so once the exchange tank is at target temp, the Tstat will let the engine cool itself if needed. Looking at the numbers, this should be an extremely efficient way to get electric and heat.

I am aware the diesel won't survive forever being run for probably 300hrs/mo, but we intend to either have parts or a spare engine on hand.

The issue is I need a battery bank. I don't like lead acid. They don't like being cycled, they need water, they are picky about temp, etc, etc. I am looking for options here??? Ideally I would like to find a 3 phase inverter??? I figure this probably won't happen or at least not cheap enough but i think the 3ph inverter could be easier to implement and allow small 3ph loads to be ran or idled. In many cases if a machine is being setup, it will need to be turned on but not actually doing work and can take many hours. Those loads are primarily just a computer, fans, and in some cases the hydraulic power units but they don't pull that much when not being exercised.
The single most reliable and long-life battery type is nickel-iron - NiFe cells.

Manufacturers typically rate them for 30 - 50 year life but in practice there are cells over 100 years old still in use. They simply do not degrade, the only "wear" factor is that the electrolyte gradually reacts with carbon dioxide in the air and so needs changing every few years.
The cells/plates do not wear out or break down, like lead-acid.

The downside is that they are big and heavy, but that's generally not a problem for an industrial site.
There has to be a reason they are not used for green energy tech. I am guessing extremely low energy density and very high costs? I am fine with big and heavy if I can bank enough power cheap enough.
Jezus, the price is right up there!!! lol Being so simple and green, are they something that could be built here? Since I need quite a bank, I can certainly entertain some fab work here. I think I recall these being the railroad battery go-to. They don't fail. I do like that. Not sure how the discharge curve looks yet though.
I don't know much about them but some companies are selling salt-water batteries, primarily for solar energy storage, and they may be worth a look.
There is plenty of info online about the construction of NiFe cells, but actually doing it properly is very difficult; one of the electrodes is made of compressed powder in a mesh sleeve, if I remember correctly - and getting the materials is not easy.

There is outline info on the wikipedia page:–iron_battery

It may work out if you are building a large number of them.

Or, there are factories in China that sell them direct, far cheaper than US or European distributors resell for.

Other types mixed in this page but you get the idea:

One appears to be 80AH cells for $10 each in quantities.
Very interesting! While not enthused to consider Asian batteries, that does look like a reasonable option! However, I am wondering about recourse in a purchase like that? While I don't mind buying a $10 flashlight from China, Spending potentially thousands on batteries worries me a lot. Returning them is not really an option financially and there is no way to vet the company is there?

I am really interested! I think I will build an out building for the generator anyway, and that building should work really well to serve as storage for the battery bank. I just don't want them in my main building for safety reasons.

I am wondering you guy's opinions on the NiFe technology vs other, newer tech such as Li-ion? I know the power density is MUCH higher and the price seems to be coming down. Maybe an Asian Li battery is a bad idea as it might have very limited life or performance? Seems they are still trying to optimize that technology in terms of capacity and life.

There was a time when I was looking at doing this years ago and appeared only flow battery tech could get me the capacity needed and flow batteries are VERY expensive currently. Personally, I feel they have the most promise of any battery tech for home use because the electrolyte is really separate from the "reactor", the electrolyte is in some cases near infinite cycle life, and the reactors can be replaced. As well, the system can scale pretty easily.

But.... to return to reality, I am still running numbers on what I need storage wise but it appears around 100kwh would be a reasonable target.
I believe what you need to do, for starters, on a clean sheet of paper, using a pencil and large eraser is define your needs allowing an adequate buffer margin. The facility I worked at had a large pump house. The pumps ran on 4160 VAC. A sub station supplying the pumps blew up. The park brought in a flatbed generator which ran 24/7 for over a week. The fuel tank was 1500 gallons and every other day a tank truck came in and put 1000 gallons of diesel fuel in that beast. That was every 48 hours. I have no clue what the capacity of the generator was but fuel consumption was 500 US gallons every 24 hours. So all I am saying is yes, I see your plan as working but depending on power needs this can get real expensive real fast. You also need to define the 3 phase inverter power. Once you have all of that compiled then start seeing which battery solution is viable. Once you have that then you start looking at a charging system to maintain the battery bank.

Electric car battery packs range between 24kWh battery packs to 30kWh battery packs. The cost has actually come down $1,000 per kWh to around $200 per kWh which is a nice improvement. They use a variety of battery chemistry, Google will likely give some results and cost. EV batteries average $3,000 to $6,000. When you start getting into the higher power demands the battery voltage is increased so the current can be reduced. Anyway, I would start by looking at EV (Electric Vehicle) battery packs, especially if you want local recourse if one fails.

I was following Aquion's work for some time. He has the "battery" figured out but what they call "cheap" is still more expensive than even Li. They did not expect Li to progress so rapidly and the over funded company went bankrupt and bought by China (super duper!)

I think it will take someone that does not think in the way of immediately going public and over funding a company to a point failure is certain. I am confident the saltwater tech could have been developed for a touch less than $130M.

We have actually been working on green technology product development so this unusual step to power a building could prove a really good excuse AND proving ground for our own product. We really want to put a few things out that do not cost so much that they don't make financial sense.
Ron, I got a good laugh about the big

This is not really a shoot from the hip idea but there are a lot of variables so nothing is perfectly set in stone. Somethings just won't be feasible to run on battery power, only when the genny is running. It is sort of a guesstimation on what will run and for how long. Harder loads are what are being considered. SImple loads such as a PC are easy and linear. Running a 1HP water pump is more tricky because then you have to predict the water use. The main factor seems to be lighting. The shop will likely have 4-5kw just in lighting! I have to predict if those will need to ever run on battery power and for how long. I can also configure them to run at 50% power. I can also install minimum lighting for this as well. This is based on already owned T5 fluorescent, but LED might be a better option right out of the gate!

As far as generator fuel costs, we have run a LOT of numbers and believe it or not, if we do this right, we can actually beat the grid price! The scenario you mention running water pumps is likely close to 1 megawatt. If you ran the numbers for the cost of 500gal/day of diesel, it is likely the price difference to the grid power probably was not much different. Right now, comparing kwh to kwh, our electric grid power costs $.14/kwh, and nat gas is $.02/kwh. We are working hard to get nat gas on site. Propane will likely turn out a better solution than diesel in this size, but nat gas trumps all by a large margin! The reason to consider nat gas is only because due to all the oil drillng on our area, they hit a LOT of gas pockets and I don't predict we will see a major change in gas price for some time. Enough time for us to look at more efficient strategies. Converting gas to electric is very poor convertion efficiency. In the order of 20%!!! that leaves serious room for improvement.
Last edited:
Here's an interesting chart that shows approximate fuel consumption of a Diesel generator for various power outputs.
So, for example, if you were running at 100kW, the fuel consumption is about 8 gal/hr.
For a wholesale Diesel fuel price of $2.50/gal, that makes the price of the electricity about 8/100 * 2.50 = $0.20 per kilowatt/hr.
Certainly more than most utility prices, but likely tolerable for most commercial backup requirements.
Crutschow, based on that chart and running offroad diesel that is 7.4* $2.25(diesel cost)/100(kw) = $.167/khw

Here is where the math gets interesting, diesel has about 139kbtu/gal, nat gas costs $.72/therm which is 100kbtu. This puts the equivalent price for the same btus as diesel right at $1.00. The prime mover won't be quite as efficient as a diesel engine BUT a whole lot cheaper to run so running on Nat gas will EASILY be half the cost. Now you are in there around $.08/kwh!!!
Last edited:
Thought you would get a kick out of the sharp pencil with a large eraser. Our facility had a 100 KW unit for emergency backup power which ran on natural gas (480 volt 3 phase delta). Right as I retired we were going to a 200 KW unit and I was happy to be leaving and one of my happiest days was when we got a go ahead to hire a facilities engineer.

Here at home on a much smaller tiny scale I just replaced our emergency backup house generator. We had a 12 KW and went to a 16 KW. Natural gas fueled since we have natural gas at the residence. Even the old 12 KW was severe overkill. I can't make my own electricity as cheap as I can buy it from the electric company. Then when I figure in needed periodic maintenance the cost further increases. Here are my little tiny numbers:

8 KW Load is 218 cubic ft. hour.
16 KW Load is 309 cubic ft. hour.

You have yourself one heck of an interesting project. You not only need to figure fuel and maintenance cost today but speculate on where the fossil fuel market will be tomorrow. Wishing you the best on this project.

8 KW Load is 218 cubic ft. hour.
16 KW Load is 309 cubic ft. hour.


That appears to be $.19/kwh and $.13/kwh respectively. Bigger engines are just more efficient so actually running a bigger shop with higher loads will be a bit better. I would not say it is really "ideal" and I am not enthused to have to deal with it but the alternatives are far more expensive. 6 figures to bring high power 3phase means I can buy and repair a lot of generators. As well, once you consider the cost of bringing electric, then you have to run the numbers against their rates and it will just take a LOT of years to see much advantage.

I won't bore you with what we are hoping to see and do down the road but I do think nat gas price will remain fairly stable. There are a lot of reasons for it but I guess even if that happens, it would have to go pretty wild to hurt us too bad.

For now we just need to figure out the batteries.
6 figures to bring high power 3phase means I can buy and repair a lot of generators. As well, once you consider the cost of bringing electric, then you have to run the numbers against their rates and it will just take a LOT of years to see much advantage.
Oh I have seen that. Fellow I know owns a rifle range and manufacturing facility in a very rural setting. Only one single wire out there. He was using motor generator converters to power his lathes and milling machinery. Maybe 25 years ago he was looking at about 100K just to get 3 phase to the end of the drive. Apparently enough demand existed so the power company ran 3 phase down the road and today the machine shop has 3 phase power. He still had to lay out a chunk just for the poles on his property.

Again, really cool and challenging project and I wish you the best through all of it and also success.

Have I missed something here?, it would cost too much to run electricity to the property (fair enough), but it's viable to run gas to the property?, or are you talking about trucking gas (or indeed diesel) to the property long term.

A friend of mine bought a fairly big generator from an auction a number of years back, I can't remember the size but it was pretty substantial, and was a two stroke diesel. It came from an American Airforce base in the UK - and was presumably sized to power the entire base?. I remember at the time we decided it was big enough to power the town we were in, about 15,000 people.

He sold it quite easily, and made a decent profit.
Well.... apparently the electric coop has zero interest or incentive to subsidize the electric build costs. All other providers will do this and we have another facility where we got 600A of 480V 3P installed FREE, and we use it! The coop is apparently NOT in the business to make money. Gas company on the other hand is examining their angle and they are a 1/2mi away and considering routing us gas free of charge. Not only does do we like free, but the cost of gas per btu has always been substantially better than electric.

We are looking closely at alternative solutions down the road using gas such as a Capstone micro turbine, or even a Bloom fuel cell. It has to make sense over a standard generator though. Green tech is evolving insanely fast right now. PV panels are getting cheap, inverters are everywhere, etc, etc.

We have some serious loads which means our options are give all our money to the electric company, or consider "investing" in our future and at least own what we pay for. It is a bold play, I realize. But we have been looking at this for many years now.
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips