50 Powerpack scenario

[deonb]

I have a family member who is running an HOA in South Africa, consisting out of around 800 fairly upscale homes.

As some of you may know from the New York Times article over the weekend, South Africa has had rolling blackouts (load shedding) for the last few years, and is expected to have that until the end of the decade. The blackouts last from 2 to 4 hours every day or two.

A lot of people are now turning to generators and individual solar/battery solutions to augment the grid, but we were wondering what the economics would be of doing a Powerpack solution instead for this whole estate. So I want to help put a slide-deck together for the HOA. (This is just for them to justify the time & cost of getting an Engineer out to do a formal assessment, not for anything more than that).

Can someone comment on whether my assumptions are reasonable?


I figured to sustain 800 houses for 2 to 4 hours you would need around 2 MW of peak supply capability, and around 4 MWh of stored energy. I'm using the 400 house/MW model, which is probably overkill. However, load shedding happens during peak times, so that would need to be considered.

To store 4 MWh of energy, at 80% efficiency, one would need 50 Powerpacks.

That would work out to be around $1.25m. Add ~$50k for shipping + insurance, and you get ~$1.3m.

Then for the inverters I haven't seen a price for a 2 MW inverter, but I can see a 1 MW inverter here for $250k. I suppose they'll need 2 of them, so $0.5m? Is this a reasonable quote? I've seen inverters as high as $150k for 100 kW (it would be scary if I need to budget $3m for the inverters!)

Let's add $250k for taxes, $200k for installation, and $150k for unforeseeables (e.g. "convincing" the utility to do proper billing for this), and this looks like to be around a $2.4m project.

There is a lot of empty land available on the estate to host these, as long as the Powerpacks don't need to be indoors (I don't think they do).

Did I miss anything else?


If this is accurate, it would mean around a $3000 one-time levy per house, which should be a fairly easy sell to the HOA board.

 

[MSEV]

Fascinating project, wish I could help you. Interesting to follow...

 

[deonb]

Actually, this is one of those no-brainer things that Tesla & Elon can solve at the government level... in theory.

South Africa's GDP is currently depressed by about 2% due to the rolling blackouts. That's $35b over 5 years.

Load shedding can be up to 4 GW, and can last for 18 hours/day, rolled over 9 regions. So the power company (Eskom) is short around 72 GWh per day when they need to shed. (This representing 10% of the total grid capacity of the day.)

Let's say the government spends that projected $35bn GDP lost to provide a grid-attached battery backup system instead. (Keep in mind that the budget for the next 2 planned power stations is expected to be around $83bn and will likely not be fully operational for the next 5 years). Now let's say of that $35bn, $25b of it is Powerpacks, it would equal 1m power packs, providing 100 GWh of storage, solving the problem, and give a little reserve.

Even better, once the under-construction coal plants (expected to provide 16.3 GW), are finished you have yourself a nice 100 GWh spare battery array in place that can support a large solar operation for future expansion. But probably more likely though is that once the new plants are up, older plants will start acting up more, so you'll still need storage for some time to come.

Unfortunately, there is just no way Tesla can manufacture 1m power packs in time for this to matter... it will be faster to just finish construction on the new coal plants under construction. However, I'm sure an opportunity like this will come up again at some point in the future, and I hope Tesla is ready to act then. (Zimbabwe actually has the same problem, but they can't afford to fix it at all).


An interesting titbit I found during research: One of the load sheddings needed to take place because it rained more than expected, and the coal got wet. So that's apparently a thing...

 

[Electric700]

Tesla installed a 100 kW system (I believe 1 Powerpack) at the University of South Florida St. Petersburg, and had the launch event two months back. May be you could propose something similar. Each Powerpack might be expandable to more than 100 kW, so fewer than 50 Powerpacks could be a possibility in such a case.

Link: Solar Launch Event at USFSP.

There's a form for Tesla Energy that you can access at the bottom of this page: Tesla Powerwall.

 

[richkae]

Can you get real energy use numbers from the HOA ( from previous electric bills)?

You might get more bang for your buck if you budgeted a portion of the money into efficiency instead of a larger battery.
Efficiency may be a lot cheaper than battery. LEDs are really cheap now.

 

[deonb]

Can you get real energy use numbers from the HOA ( from previous electric bills)?

You might get more bang for your buck if you budgeted a portion of the money into efficiency instead of a larger battery.
Efficiency may be a lot cheaper than battery. LEDs are really cheap now.

Most people in there use pre-paid electricity, so there is no central meters at the estate level. For the HOA to get access to it requires going around to everybody. Based on a few inputs though, it's quite possible the peak draw may be as low as 1 MW, but I'll do the math for 2 MW anyway.

The homes are mostly built using 12V or 24V LED lighting (this was before 240V LED's became available).

Temperature is modest (-ish) throughout the year, so few people have air conditioners or central heating. Things are pretty much as efficient as it will get - electricity has been expensive for a while, which caused people to optimize already.

So the main draws on the power grid are electrical hot water heaters and stoves/ovens, with some strain from refrigerators/freezers. These far overshadow any further gains that you can make elsewhere. Unfortunately there is no access to gas, and no real possibility for it, so not much can be done about it.

The one possibility would be to however control the hot water heaters centrally, but that would probably add more cost to the project than the expense in powerpacks & inverters to just drive them. It also requires more of a commitment from people, and much bigger organization & planning.

- - - Updated - - -

The one thing I'd like to add - I don't know if the Powerpacks and Inverters are really all you need? There isn't solar input, so it's draw from the grid when available, and do a UPS switchover when not.

Most equipment planning layouts I've seen has more to do with a solar scenario rather than pure battery.

 

[electracity]

I'm sure there are microgrid projects with solar and battery happening in SA. A subdivision is a natural microgrid. What is the cost of electricity? I doubt the subdivision as a whole would be allowed to add central batteries without solar. It just makes the rolling blackout situation worse for everyone else..

 

[bwa]

An interesting titbit I found during research: One of the load sheddings needed to take place because it rained more than expected, and the coal got wet. So that's apparently a thing...

Yes it is. I have intimate experience with that sitting on top of a coal mill on graveyard shift inside of a cement factory in California. They never invested in much weather infrastructure (to handle rain) because it would not have been cost efficient since it almost never rains in California. Thus during an unusually wet year I spent a month with a compressed air gun and various brushes to stop the coal chute from gunking up with wet coal, stopping the flow of fuel, stopping the factory and costing millions. It was boring work, and I had a lot of time to sit and think about energy. But yes, of course it is a thing, I say, but that's only because I did that. If you think about it, it's not too surprising: how are you going to keep mountains of coal dry during rain? Whatever you do, it is expensive and requires its own maintenance, cleaning, etc., such as covers 300' in the air getting clogged with dust. It requires real engineering and cost whatever you do, so no quick fixes. Coal is dirty just like a lot of things. More dirty, even. Hell, I'd go as far as to say it's dirtier than enough solar panels to cover all our energy needs. Just think about that: rain makes dirty energy clog. Rain comes from solar energy itself, not surprisingly. I could literally sit there for a whole shift mulling over it in my mind as the mill mulled the coal to explosive dust beneath me.

 

[bwa]

I doubt the subdivision as a whole would be allowed to add central batteries without solar. It just makes the rolling blackout situation worse for everyone else..

Exactly my thought. Why not get enough solar panels to supply most your own energy needs as well as the batteries? Dark day planning would be nontrivial, but I think there must be a way.

Since hot water is a big user of energy in your scenario, augmenting solar with solar water heating would defray a lot of conversion costs (less electric solar, less battery, less inverter). A dark day study would have to take all of this into account, using cost offsets from bright days.

Someone could explain how to plan for a dark day: do you pre-charge the batteries before the dark day? That requires a lot of weather prediction. How do you handle forecast errors? How does the controller manage this? Do you oversize for errors and always keep two days reserve at 4PM but if it's not enough then you charge that following night from exterior supplier grid up to one day reserve, regaining your one day reserve and hoping light charges to two days the following day and repeating cycle, thus inverters and panels are sized for expected use but batteries for two days? Ok, so the actual count wouldn't be at 4PM, but at 10PM, which if batteries are less than one day reserve, they charge from grid up to one day reserve then stop charging. They hope to charge to their full two day reserve mid day but always have at least one day to get to 10PM. Aha, I answered my own question.

 

[deonb]

I'm sure there are microgrid projects with solar and battery happening in SA. A subdivision is a natural microgrid. What is the cost of electricity?

It's not directly cost-economical. Cost of electricity is around US 14c / kWh.

Besides, you're talking about a capital layout that's many orders of magnitude bigger. Now you need to not just be able to handle 2 hours of storage during load shedding, but 72 hours of storage during dark days.

I doubt the subdivision as a whole would be allowed to add central batteries without solar. It just makes the rolling blackout situation worse for everyone else..

I can't see how it would make it worse. It shifts demand from peak hours to off peak hours. Utilities generally have about a 50% difference in demand between peak and off peak, and they'll welcome any solution that can shift demand like that.

 

[johnbarbour]

deonb,

How is your project going?

I'm doing a similar assessment, with a bit more detail for the Atlanta area. Would love to see what else you have found on the inverter situation.

 

[Cyberax]

So the main draws on the power grid are electrical hot water heaters and stoves/ovens, with some strain from refrigerators/freezers. These far overshadow any further gains that you can make elsewhere. Unfortunately there is no access to gas, and no real possibility for it, so not much can be done about it.

Have you thought about buying liquefied petroleum gas? It can be safely stored in a central (preferably underground) tank and distributed to individual houses using low-pressure tubing. LPG systems are completely passive and are pretty much maintenance free after the initial installation. You'll just have to occasionally buy a tanker truck-load of LPG to refill the central tank.

Such a system will allow you to shed quite a bit of electrical load used for water heating and stoves, probably saving you a lot of money.

I've seen such system in action in Russia (in a Siberian village), it was also used for electricity generation with a small gas turbine.

 

[Kurt Foster]

H

I have a family member who is running an HOA in South Africa, consisting out of around 800 fairly upscale homes.

As some of you may know from the New York Times article over the weekend, South Africa has had rolling blackouts (load shedding) for the last few years, and is expected to have that until the end of the decade. The blackouts last from 2 to 4 hours every day or two.

A lot of people are now turning to generators and individual solar/battery solutions to augment the grid, but we were wondering what the economics would be of doing a Powerpack solution instead for this whole estate. So I want to help put a slide-deck together for the HOA. (This is just for them to justify the time & cost of getting an Engineer out to do a formal assessment, not for anything more than that).

Can someone comment on whether my assumptions are reasonable?


I figured to sustain 800 houses for 2 to 4 hours you would need around 2 MW of peak supply capability, and around 4 MWh of stored energy. I'm using the 400 house/MW model, which is probably overkill. However, load shedding happens during peak times, so that would need to be considered.

To store 4 MWh of energy, at 80% efficiency, one would need 50 Powerpacks.

That would work out to be around $1.25m. Add ~$50k for shipping + insurance, and you get ~$1.3m.

Then for the inverters I haven't seen a price for a 2 MW inverter, but I can see a 1 MW inverter here for $250k. I suppose they'll need 2 of them, so $0.5m? Is this a reasonable quote? I've seen inverters as high as $150k for 100 kW (it would be scary if I need to budget $3m for the inverters!)

Let's add $250k for taxes, $200k for installation, and $150k for unforeseeables (e.g. "convincing" the utility to do proper billing for this), and this looks like to be around a $2.4m project.

There is a lot of empty land available on the estate to host these, as long as the Powerpacks don't need to be indoors (I don't think they do).

Did I miss anything else?


If this is accurate, it would mean around a $3000 one-time levy per house, which should be a fairly easy sell to the HOA board.

Heck if rolling blackouts were more often worldwide, we would evolve to use solar panels... I think