Building a massive off-grid inner city system

[ecoco]

Hi all, I've been a fan of will's channel for a while now and I think I'm finally ready to start planning my solar system, but I have a questions around feasibility that I hope someone could help me answer.

A little background first: I live in Indonesia, so we get plenty of sun. But during the rainy season we obviously less. I can't exactly say how much less but it is pretty rare that we have a full day or two of rain. It comes and goes pretty quick. That being said, the contractors in Indonesia I've been in touch with use the following equipment:

- batteries: http://www.pylontech.com.cn/ or https://www.todoensolar.com/Pylontech-lithium-battery-48V-35kw-/-h-US3000s
- Inverter, monitoring system etc: https://www.victronenergy.com/
- Panels: (not sure yet)

The primary purpose of this system is to be able to power a central AC unit that will cool the entire house all the time. I'm assuming we will install a 5 ton unit (60,000 BTU I think) along with everything else. I've made a short of list of my power needs (see image) but there might be additional items so it would be best to oversize the system. I think it's safe to assume I'll need a 86kwh system.



My question is, what is a good assumption in terms of power consumption for the AC unit? I want the unit to run 24/7 but I doubt it will be drawing full power all the time, so how do I go about reasonably estimating the power usage? I'd like the batteries to be able to support 2-3 days of minimal sun, and be able to charge fully in a typical day (is that even possible?).

I plan on tying to the grid (and generator eventually), but only as a backup source of power in case the batteries need topping up or we want to do maintenance on the system. I can sell power back to the government here, but I think they only provide discounts on the normal power usage... It would be impossible to demand a check in the mail for the power I actually sell back

Any and all advice is much appreciated

 

[RV10flyer]

I would install at least a 20 SEER Trane or American Standard HP system with an inverter compressor/fan. No high current startups with this system.

I think 82 kWh is a little high. I have 5 tons total with three systems for 2700 sqft. In my all electric house, I consume about 40-60 kWh/per day. I’m building a 20.6 kw array, two Sol-Ark 12K and 107 kWh LFE battery bank.

 

[ecoco]

I would install at least a 20 SEER Trane or American Standard HP system with an inverter compressor/fan. No high current startups with this system.

Ah I see. I should go with an inverter system then. Makes sense, but I doubt I could find this brand in Indonesia.

Generally available brands are:
- Daikin
- Sharp
- Panasonic
- Gree
- Samsung

 

[acdoctor]

I think you are building a new house. Has someone done a heat loss heat gain calculation on it. With the numbers from it and the efficiency of the proposed equipment you should be able to calculate the energy consumption fairly close. Be sure to look at the efficiency of the exact equipment. Most vary in an efficiency line. Usually conventional lines a 3 ton is the one they post. The smaller ones are more and the larger ones are less efficient. I would look at two systems in Lou of one if you need a five ton. Duck losses will be less if nothing else.

 

[chrisski]

Also recommend getting some numbers for how much solar you can get in your area. If you go with the link in my site, it will help give you an idea. THat site is supposed to also take into account weather. PV Watts is another site to look at.

Here’s a graphic for Bandung 10 kw of panels, 10 kwh batteries, and 8kw used per day:



You can adjust your numbers as needed for battery bank size, panel size, and max power usage. My guess is Rainy season is June and July.

 

[ecoco]

I think you are building a new house. Has someone done a heat loss heat gain calculation on it. With the numbers from it and the efficiency of the proposed equipment you should be able to calculate the energy consumption fairly close. Be sure to look at the efficiency of the exact equipment. Most vary in an efficiency line. Usually conventional lines a 3 ton is the one they post. The smaller ones are more and the larger ones are less efficient. I would look at two systems in Lou of one if you need a five ton. Duck losses will be less if nothing else.

Good point! I never thought about using 2 systems and yes, this will be a new house. I don't have the drawings done yet either so it's hard to dictate how many BTUs I actually need right now. I'm just trying to figure out the panel footprint so that i can relay this information to my architect.

Also recommend getting some numbers for how much solar you can get in your area. If you go with the link in my site, it will help give you an idea. THat site is supposed to also take into account weather. PV Watts is another site to look at.

Here’s a graphic for Bandung 10 kw of panels, 10 kwh batteries, and 8kw used per day:
View attachment 76952


You can adjust your numbers as needed for battery bank size, panel size, and max power usage. My guess is Rainy season is June and July.

Thanks for this! Im not in bandung but its fairly close. Rainy season for us is winter for you folks in the states

 

[Hedges]

Would it be practical to build ice in an insulated tank during the day, and use that for chilling at night?
H2O is cheaper than Li of Pb. So is Si.

I'd rather use a variable-speed refrigeration compressor, controlled to keep battery at float voltage, so 100% of surplus PV generated power is used to freeze water. Then battery only has to be big enough for nighttime loads including air handler and/or antifreeze/water circulator pump.

 

[ecoco]

Would it be practical to build ice in an insulated tank during the day, and use that for chilling at night?
H2O is cheaper than Li of Pb. So is Si.

I'd rather use a variable-speed refrigeration compressor, controlled to keep battery at float voltage, so 100% of surplus PV generated power is used to freeze water. Then battery only has to be big enough for nighttime loads including air handler and/or antifreeze/water circulator pump.

I've never heard of using ice to cool a house, but from what i know, ice machines use huge amperage. I might stick to using smaller, separate AC units as acdoctor mentioned. That would give me more control on the duct runs and actually minimize the distance the cool air would need to travel.

I'd rather use a variable-speed refrigeration compressor

I'll keep this in mind when sourcing appliances thanks!

 

[noenegdod]

Would it be practical to build ice in an insulated tank during the day, and use that for chilling at night?
H2O is cheaper than Li of Pb. So is Si.

I'd rather use a variable-speed refrigeration compressor, controlled to keep battery at float voltage, so 100% of surplus PV generated power is used to freeze water. Then battery only has to be big enough for nighttime loads including air handler and/or antifreeze/water circulator pump.

The compressor has to work a lot harder to drop the additional 25-30 degrees C so the daily energy consumption to cool the house during the day and make ice to cool after dark would be substantially higher than it would be to to just cool the house with AC. It would require more panels, less batteries, an additional piece of equipment to make ice. Not sure it would be worth it.

 

[noenegdod]

I've never heard of using ice to cool a house, but from what i know, ice machines use huge amperage. I might stick to using smaller, separate AC units as acdoctor mentioned. That would give me more control on the duct runs and actually minimize the distance the cool air would need to travel.

I cool my house with one 2.5ton unit. It is not big enough to keep it at 20C on the hottest of days, it will warm up to 23 then. During the summer it basically turns on in June and doesnt turn off very often until September. I say this because if you use 2 units in a zone manner they will tend to cycle on and off. If you use 2 units but have them both cooling the entire house, one will turn on and just run, if it can not keep up the second one will start and help out. At night, the one will just run as required. Just some food for thought.

 

[ecoco]

I usually set my AC here to 25C. Any lower and it's actually too cold for most Indonesians. The key is moisture. Once the moisture hits around 50%-60%, it's very comfortable at 25C. Since the house will be conditioned continuously, i don't think the ambient temp needs to be much lower than that.

I plan on building a well sealed house, but not insulated.. so maintaining 25C shouldn't be too hard. The average temp here is about 30C. I can hear my current split AC compressor turn on for about 10 minutes at a time every hour or so at night - and the unit is set to 26C. So I think power consumption will drop even more at night.

 

[Hedges]

I've never heard of using ice to cool a house, but from what i know, ice machines use huge amperage. I might stick to using smaller, separate AC units as acdoctor mentioned. That would give me more control on the duct runs and actually minimize the distance the cool air would need to travel.

This has been used in industrial applications, taking advantage of low utility rates at night and reducing power draw during the day.

The compressor has to work a lot harder to drop the additional 25-30 degrees C so the daily energy consumption to cool the house during the day and make ice to cool after dark would be substantially higher than it would be to to just cool the house with AC. It would require more panels, less batteries, an additional piece of equipment to make ice. Not sure it would be worth it.

True, making 0 degrees C vs. 25 degrees C is harder. It is also dumping heat into the hotter daytime air vs. cooler nighttime. Unless it is water source. Since this is "inner city" less likely to have surface water or ground water as a heat sink.

Not sure how much extra power wasted for these two effects. Guys who know HVAC calculations could work it out.

I figure PV panels cost US$0.025/kWh amortized over a decade. Batteries cost $0.05 to $0.50/kWh over their lifetime. Assuming they last their projected lifetime. PV panels can last 25 to 40 years (barring quality issues causing excessive degradation), so they just keep getting cheaper.

Compared to $0.50/kWh batteries of course I'd take the thermal inefficiency hit and make ice. If you can do batteries for $0.05/kWh, off the shelf energy system and HVAC components would be attractive.

 

[12VoltInstalls]

never heard of using ice to cool a house, but from what i know, ice machines use huge amperage

Ice is cold. You remove heat from water and ‘waste’ the heat out of water and release the heat to the exterior.
Just like AC cooling does. But the ice suggestion stores this low-energy heat for later use. Since the energy (in moles) to maintain X interior temperature is ‘static’ in this scenario, this might be a worthwhile arrangement to explore. Basically you’re storing the energy in ice rather than batteries.

control on the duct runs and actually minimize the distance the cool air would need to travel.

Other than maybe fan watts to move the cooling medium (air) any duct “loss” is still effectively cooling the cubic footage of the interior.

plan on building a well sealed house, but not insulated

Insulation would do three things for you:
1) realize a significant reduction in the energy required to maintain temperature
2) the needed equipment size/capacity will be smaller
3) the cooled space will have a more stable temperature
I think you really need to insulate.

Not sure how much extra power wasted for these two effects. Guys who know HVAC calculations could work it out.

I’m no engineer but I’ve managed projects with central and split air from 5KSF to 45KSF. With pre-cooling I think will “lose” on the initial 24-hr cycle BUT if the cooling/stored low temp mass is within the cooled space any ‘losses’ are effectively losing to and creating the desired climate controlled environment (not in a basement or outbuilding in other words). Your idea has merit.

I'd take the thermal inefficiency hit and make ice.

The ‘practical’ problems to solve are many but it’s doable. In this situation I think the scale of residential might make the engineering and install cost prohibitive vs manufactured mini splits. Secondly, you still have to have some central or split unit to both distribute the low-temp energy and dehumidify it.

But I like your idea. Like the tulikivi (sp?) of the 1400s for heat, the evaporative central stone cooling of the southwest US from a few hundred years ago, or today where wood furnaces are used to heat water that is stored in 6000 gallon water tanks; this essentially takes an energy, stores it, and releases it later over time.

 

[ecoco]

Insulation would do three things for you:

I know it sounds like a good idea to insulate, but we're talking about vastly different climate zones. Insulation is great for saving the cost of heating a house, but cooling is no where near as expensive. There are also issues with finding the talent who can to build a brick insulated cavity wall in the country. We don't use lumber so putting up insulation is far more difficult and creating a sealed space is that takes advantage of the insulation not so straight forward.

Generally, we just try to keep the heat from getting at all. The sun is harsh during the day, but the moisture lingers. That moisture, if trapped in the space where insulation would go could wreck havoc on anything that might corrode, or rot. We also get torrential downfalls, so if the house does not drain well, you'll be left with water finding it's way into every nook and cranny.

As much as I appreciate the brainstorming on cooling efficient solutions, I think we're looking track of my initial question. I really just need to know how to estimate the power usage of a large AC unit, or a few smaller ones. I want to build a system where I just don't have to think about usage. Where I can easily let the AC run 24/7, even if its set to minimal settings just to keep the house comfortable.

I'm trying to budget around 60k for this system - is that doable?

Can someone help with this?

 

[acdoctor]

Do you have a set of prints and a list of construction components and methods? If you do you can get a load calculation done. If not you can find someone with a very similar house. Anything else is a guess. Looking back through post I don’t even see square footage.

 

[ecoco]

Do you have a set of prints and a list of construction components and methods? If you do you can get a load calculation done. If not you can find someone with a very similar house. Anything else is a guess. Looking back through post I don’t even see square footage.

Not yet, the architect has barely started. Maybe can just calculate the load of a typical 60k BTU AC unit running 24/7 first?

 

[Nobodybusiness]

I would install at least a 20 SEER Trane or American Standard HP system with an inverter compressor/fan. No high current startups with this system.

I think 82 kWh is a little high. I have 5 tons total with three systems for 2700 sqft. In my all electric house, I consume about 40-60 kWh/per day. I’m building a 20.6 kw array, two Sol-Ark 12K and 107 kWh LFE battery bank.

Same size and components of my system.

 

[acdoctor]

Typically here recommended size is 20 to 30% larger than calculations.

 

[ecoco]

Same size and components of my system.

Oh wow, sorry how did i miss this. This is a great ballpark figure. I'll take it to my solar installer when they get back from holidays (cause here, they just try to sell you anything).