RV Air Conditioner and Solar - Again

[DanConnors]

Sorry for the long tale, but I think the context matters here. In 2007, our home burned down due to wildfire and we rebuilt in the same place. Some people lost their pets because the pets were in their home when the roads were closed and the owners could not return and retrieve them. We rebuilt on the same site but determined whenever we left, the dogs would be with us.

We live in the mountains east of San Diego at 2750'. The sun is usually out and it is intense. It's 18 miles to the closest grocery store, so we have installed a 110V smallish refrigerator in a Transit Connect. It is powered by an inverter powered by three 110AH deep cycle flooded batteries and recharged using shore power. We also have a Fantastic vent fan in the ceiling and two dog cages that heat/cool using 12v power. This combination works fine until the outside air temperature exceeds 80 degrees. Above that, we use our Roadtrek Class B motorhome, which has a generator, air conditioner, and a gas hungry engine.

The Transit Connect needs a new transmission, and we are looking to replace both the Transit Connect and the Roadtrek with another vehicle that has Air Conditioning independent of the vehicle's system powered by the motor (so we have air conditioning with the car engine off). There have been several example of how much solar/battery power you would need to power air conditioning here on the forum, but the math seems to assume the air conditioning is in use when the solar panels are not working. I suspect that is a rare case in San Diego.

So here's the question. I can find a 5000 BTU 110v window A/c that consumes 550 watts. I can find numerous flexible 100w solar panels. I see that solar panels are less efficient when hot, so I guessed 80w out of a 100w panel. Using six panels, that would produce about 480 watts of power--the additional 70 watts would come out of the batteries.110AH x 3 = 330AH *12 = 3960 watt hours, divided by 70 watts (the shortfall) = 56 hours. Since the intense heat rarely exceeds 8 hours, there seems to be no problem here--which is never the answer I see to the question, "Can I power an air conditioner with solar?"

So, there is a pretty good chance I misinterpreted something. Will you please provide your thoughts/corrections and especially ideas on whether it would be smarter to go to 24v (bearing in mind I still have 12v things to power).

Thanks!

[max2k]

Sorry for the long tale, but I think the context matters here. In 2007, our home burned down due to wildfire and we rebuilt in the same place. Some people lost their pets because the pets were in their home when the roads were closed and the owners could not return and retrieve them. We rebuilt on the same site but determined whenever we left, the dogs would be with us.

We live in the mountains east of San Diego at 2750'. The sun is usually out and it is intense. It's 18 miles to the closest grocery store, so we have installed a 110V smallish refrigerator in a Transit Connect. It is powered by an inverter powered by three 110AH deep cycle flooded batteries and recharged using shore power. We also have a Fantastic vent fan in the ceiling and two dog cages that heat/cool using 12v power. This combination works fine until the outside air temperature exceeds 80 degrees. Above that, we use our Roadtrek Class B motorhome, which has a generator, air conditioner, and a gas hungry engine.

The Transit Connect needs a new transmission, and we are looking to replace both the Transit Connect and the Roadtrek with another vehicle that has Air Conditioning independent of the vehicle's system powered by the motor (so we have air conditioning with the car engine off). There have been several example of how much solar/battery power you would need to power air conditioning here on the forum, but the math seems to assume the air conditioning is in use when the solar panels are not working. I suspect that is a rare case in San Diego.

So here's the question. I can find a 5000 BTU 110v window A/c that consumes 550 watts. I can find numerous flexible 100w solar panels. I see that solar panels are less efficient when hot, so I guessed 80w out of a 100w panel. Using six panels, that would produce about 480 watts of power--the additional 70 watts would come out of the batteries.110AH x 3 = 330AH *12 = 3960 watt hours, divided by 70 watts (the shortfall) = 56 hours. Since the intense heat rarely exceeds 8 hours, there seems to be no problem here--which is never the answer I see to the question, "Can I power an air conditioner with solar?"

So, there is a pretty good chance I misinterpreted something. Will you please provide your thoughts/corrections and especially ideas on whether it would be smarter to go to 24v (bearing in mind I still have 12v things to power).

Thanks!

problem with AC powered by solar is thermal inertia- things remain hot for 2-3 hours after sunset when solar power goes to 0. Solar 'wakes' up earlier in the morning when everything is still cool and AC is not yet needed. So batteries must be able to power AC for those 2-3 hrs at full power. Is it possible to get grid power to the house?

[SunEagle]

The biggest misconception most people have is that you won't need AC cooling when the sun isn't shining. As max2k stated the latent heat of the day is still around well after solar panels can't produce any power so some type of battery system is needed to keep the cool air equipment running.

The problem with using batteries is that you need 415 Ah 12volt battery system to safely generate 1kWh a day which is not enough to keep a 550 watt AC unit running 2 hours. So you need even more batteries or a much smaller AC unit.

[littleharbor]

So here's the question. I can find a 5000 BTU 110v window A/c that consumes 550 watts. I can find numerous flexible 100w solar panels. I see that solar panels are less efficient when hot, so I guessed 80w out of a 100w panel. Using six panels, that would produce about 480 watts of power--the additional 70 watts would come out of the batteries.110AH x 3 = 330AH *12 = 3960 watt hours, divided by 70 watts (the shortfall) = 56 hours. Since the intense heat rarely exceeds 8 hours, there seems to be no problem here--which is never the answer I see to the question, "Can I power an air conditioner with solar?"

So, there is a pretty good chance I misinterpreted something. Will you please provide your thoughts/corrections and especially ideas on whether it would be smarter to go to 24v (bearing in mind I still have 12v things to power).

Thanks!

What's going to charge your batteries when you are drawing power from them and panels are only powering the AC? Keep in mind your panels , unless on a tracker, aren't going to be putting out peak power except during mid day, and then only if oriented properly.

[DanConnors]

Can you please help me understand this: "The problem with using batteries is that you need 415 Ah 12volt battery system to safely generate 1kWh a day which is not enough to keep a 550 watt AC unit running 2 hours. So you need even more batteries or a much smaller AC unit." Am I incorrect that 415AH at 12v is 4980 watts for one hour? How did you get to that is less than 1 kWh?

I have no experience with solar, so maybe someone could come up with a guess: if I have solar panels flat on the roof of a vehicle (not tracking), and only use the vehicle from noon to 6PM, what percent of the rated output of the solar panels should I expect in San Diego (31 degrees north latitude) with an average outdoor temperature of 85 degrees? Maybe there is a lookup table somewhere for this?

My use case is that the vehicle will be parked in the sun and plugged in until noon, so batteries will be 100% charged. Intention is also to have batteries charged from car generator, but let's disregard that. Once the outside temp drops below 80 degrees, we will turn off the air conditioning and use a fan to vent.

[J.P.M.]

Can you please help me understand this: "The problem with using batteries is that you need 415 Ah 12volt battery system to safely generate 1kWh a day which is not enough to keep a 550 watt AC unit running 2 hours. So you need even more batteries or a much smaller AC unit." Am I incorrect that 415AH at 12v is 4980 watts for one hour? How did you get to that is less than 1 kWh?

I have no experience with solar, so maybe someone could come up with a guess: if I have solar panels flat on the roof of a vehicle (not tracking), and only use the vehicle from noon to 6PM, what percent of the rated output of the solar panels should I expect in San Diego (31 degrees north latitude) with an average outdoor temperature of 85 degrees? Maybe there is a lookup table somewhere for this?

My use case is that the vehicle will be parked in the sun and plugged in until noon, so batteries will be 100% charged. Intention is also to have batteries charged from car generator, but let's disregard that. Once the outside temp drops below 80 degrees, we will turn off the air conditioning and use a fan to vent.

See PVWatts for a decent modeling of long term output output of a PV system. Actual 30 day output variation from model output will be +/- ~~ 30% or so, largely dependent on weather variability - actual weather will be different than what the model uses. See the PVWatts help/info screens for particulars. Models are not predictors.

A very approx. SWAG at output in inland San Diego county is roughly from about 120 kWh/month (Jan.)to about 160 kWh/month (Aug.) per STC kW installed on a horizontal plane.

Any battery system and DC to AC schemes for running an AC air conditioner are best sized/designed taking both solar output and the load characteristics into account.

[SunEagle]

My quick calculations determined if you use a 415Ah 12volt battery you can safely get about 1000 watt hours out of it a day.

I used the formula of 415Ah x 12v = 4980 watt hours x 20% (daily discharge) = 996 watt hours. You can discharge a battery system to 25% but run the risk of over discharging it and cutting short the number of daily life cycles.

A 415Ah battery should be charged by about 41 amps which quickly calculates to 500 watts of panels using an MPPT CC. 41amps x 12v = 492 watts.

Now to get that 41 charging amps for at least 4 hours (10 am to 2 pm) you will need to have the panels pointed South directly at the sun at an angle equal to your latitude location. If the panels are lying flat on the RV roof or not oriented due South then you will generate 41 amps for less hours.

[DanConnors]

Thank you for continuing this conversation! You used the term "20% daily discharge." I interpreted what you said to mean the battery is at 80% capacity when you stop discharging it. I thought the standard without damage was 50% (no laughing, last week I thought it was 100%). I'm guessing if I used 600 watts of solar power, that would help compensate for having them flat (and would fit on the surface I have). If 492 is output from 500w of solar then 590 is output from 600w. Do you think flat mounting with usage from noon to five PM could be reasonably expected to produce 300w (50%)? If yes, then the shortfall is about 250w, requiring 250/12 = ~ 21A/hour from the battery for 5 hours. 105AH is 20% of 525 AH, so five 110AH batteries could perform for 5 hours at 20% discharge. Using the three I have (330AH, they would discharge about 30% in five hours.

Finally (I hope). Once I get the general parameters figured out, I obviously need to find a professional to design all the electrical stuff. Since this is a mobile application, can you say what kind of professional I should look for? I don't know if recommendations are allowed on this forum, but I would gladly accept any if permitted.

Thanks again!

[J.P.M.]

My quick calculations determined if you use a 415Ah 12volt battery you can safely get about 1000 watt hours out of it a day.

I used the formula of 415Ah x 12v = 4980 watt hours x 20% (daily discharge) = 996 watt hours. You can discharge a battery system to 25% but run the risk of over discharging it and cutting short the number of daily life cycles.

A 415Ah battery should be charged by about 41 amps which quickly calculates to 500 watts of panels using an MPPT CC. 41amps x 12v = 492 watts.

Now to get that 41 charging amps for at least 4 hours (10 am to 2 pm) you will need to have the panels pointed South directly at the sun at an angle equal to your latitude location. If the panels are lying flat on the RV roof or not oriented due South then you will generate 41 amps for less hours.

As for the air conditioner load, that's a problem in heat transfer. The A/C will run, probably intermittently, whenever the thermostat calls for cooling. On hot days it'll run more. Once the cooling load and its temporal characteristics are estimated - say the compressor will run 4 hours out of 8 on hot Aug. days (maybe between 1100 to 1900 hrs. or so), and maybe 1 hour out of 5 on hot days in Jan. (say between 1300 and 1800 hrs. or so) - something like that, the array can be sized to meet maybe 125 % of the worst (largest) load/solar availability ratio.

Once the cooling load is estimated, I'd divide it by a C.O.P of ~ 3 and get the electrical energy requirement with the idea for how long an A/C that consumes 500 Watts needs to run to satisfy the cooling demand at various times of the year.

The, say, 4 hour availability of solar energy in Jan. will be less than the 4 hour availability of solar energy in Aug., but the A/C load in Jan. will probably be less than the Aug. load, putting load and availability somewhat in sync. so the "worst" month may not be Jan. as is usually assumed, and, depending on array orientation, may well be in Jan., or Aug., or somewhere in between. Once that worst case daily load is guiesstimated, the battery (storage) and associated electronics can be sized taking the load, battery and motor characteristics into account, as well as how many days of storage may be required. The whole deal may well take an iteration or 2, but such things tend to converge rather quickly, and given the weather variability and other imponderables, good eyeball estimates are usually sufficient, or at least not disastrous.

[SunEagle]

Thank you for continuing this conversation! You used the term "20% daily discharge." I interpreted what you said to mean the battery is at 80% capacity when you stop discharging it. I thought the standard without damage was 50% (no laughing, last week I thought it was 100%). I'm guessing if I used 600 watts of solar power, that would help compensate for having them flat (and would fit on the surface I have). If 492 is output from 500w of solar then 590 is output from 600w. Do you think flat mounting with usage from noon to five PM could be reasonably expected to produce 300w (50%)? If yes, then the shortfall is about 250w, requiring 250/12 = ~ 21A/hour from the battery for 5 hours. 105AH is 20% of 525 AH, so five 110AH batteries could perform for 5 hours at 20% discharge. Using the three I have (330AH, they would discharge about 30% in five hours.

Finally (I hope). Once I get the general parameters figured out, I obviously need to find a professional to design all the electrical stuff. Since this is a mobile application, can you say what kind of professional I should look for? I don't know if recommendations are allowed on this forum, but I would gladly accept any if permitted.

Thanks again!

The number of cycles in the battery system lifetime varies based on the manufacturer. Most FLA type batteries show the longest lifetime by not discharging more than 20% although some indicate a limit of 25%. Since I do not know what type batteries you will be getting I made the safe assumption to use the 20% DOD although you can use more if the batteries don't suffer.

My wattage calculations are based on approximate numbers and can vary depending on the make and model of the MPPT charge controller and a perfectly sunny day. The best output from solar panels are usually between the hours of 10Am and 2Pm for most days if they are point due South and at angle equal to your latitude.

Calculating the output of X watts for Y hours at Z angle will provide you multiple different watt hour generation any day of the year. So it would be a toss up to what you will really get.

Last point I would like to make is for you not to wire batteries in parallel to increase the overall Ah rating. Paralleled wired batteries will not charge and discharge equally due to differences in the resistance of each "path". That usually results in one or more batteries doing the most work and not lasting as long as expected.

A battery system is best comprised of low voltage cells (2v, 4v or 6v) rated at the Ah needed and wired in series to create the final voltage and Ah rating of the system.

[SunEagle]

As for the air conditioner load, that's a problem in heat transfer. The A/C will run, probably intermittently, whenever the thermostat calls for cooling. On hot days it'll run more. Once the cooling load and its temporal characteristics are estimated - say the compressor will run 4 hours out of 8 on hot Aug. days (maybe between 1100 to 1900 hrs. or so), and maybe 1 hour out of 5 on hot days in Jan. (say between 1300 and 1800 hrs. or so) - something like that, the array can be sized to meet maybe 125 % of the worst (largest) load/solar availability ratio.

Once the cooling load is estimated, I'd divide it by a C.O.P of ~ 3 and get the electrical energy requirement with the idea for how long an A/C that consumes 500 Watts needs to run to satisfy the cooling demand at various times of the year.

The, say, 4 hour availability of solar energy in Jan. will be less than the 4 hour availability of solar energy in Aug., but the A/C load in Jan. will probably be less than the Aug. load, putting load and availability somewhat in sync. so the "worst" month may not be Jan. as is usually assumed, and, depending on array orientation, may well be in Jan., or Aug., or somewhere in between. Once that worst case daily load is guiesstimated, the battery (storage) and associated electronics can be sized taking the load, battery and motor characteristics into account, as well as how many days of storage may be required. The whole deal may well take an iteration or 2, but such things tend to converge rather quickly, and given the weather variability and other imponderables, good eyeball estimates are usually sufficient, or at least not disastrous.

All true but IMO it is hard to calculate exactly what would be needed for a solar / battery system for any specific day because of the actual watt hour usage needed to maintain the cooler temperature, the actual daily temperature of the air and the amount of sunlight the panels will get to power the AC system and charge the batteries.

It is all a guess and the OP needs to know some days they will be cool and other days they will be hot.

Anyone expecting a perfect system to stay cool every hot day using solar panels and a battery system is having a pipe dream.

As you have stated many times, "take what you want and scrap the rest"

[AzRoute66]

As I understand it, you wish to take day trips from your house, up to six hours, and power an AC unit that draws 550W, for those six hours only, then return home to shore power to charge your 12V 330 Ah battery bank. I'll take a stab at it, and if I have misunderstood your post perhaps you can use some of what I present as a template.

Worst case load: 6 hrs x 550W = 3300 Wh.

Solar Supply: 600W x 6 hrs x 60% = 2168 Wh
Battery Supply: 12V x 330 Ah x 50% Depth of Discharge = 1980 Wh
Total Supply: 2168 Wh + 1980 Wh = 4148 Wh, which clearly exceeds your worst case demand of 3300 Wh. I say go for it.

For the solar supply I estimated an average of 60% rated capacity for harvest from your panels. This is the WRONG number. I do not know the right number, it depends on your panel type (flexible?, sticky?, mono?, etc.) and even where/how you park the vehicle. Proper application of PVWatts, using 0 tilt and your location would be the place I would start, but if I already had a panel, I would just test it as a sanity check.

For the 50% Depth of Discharge I used 50% because it is the number I thought would receive the least amount of flak from others here. It is NOT the definitive number There are many schools of thought on this and it depends. I think 50% (or more) discharge is fine for your app because I don't think you will be doing your field trips daily, probably a couple times a week at most - and this only really applies to the summer trips.

Worst case load isn't worst case load. I suspect there is an inverter involved here and it has some loss. Air-conditioners have power factors and surges and stuff that I didn't consider. Appliances generally run lower power than their nameplate power. I doubt that it is going to run continuously for six whole hours. If it does run continuously, one has to wonder if the temp is really down to where it is comfortable for the pets.

Now that I have disclaimered this to the point of a complete non-answer, we'll let the smarter people refine it and perhaps get something usable.

[J.P.M.]

All true but IMO it is hard to calculate exactly what would be needed for a solar / battery system for any specific day because of the actual watt hour usage needed to maintain the cooler temperature, the actual daily temperature of the air and the amount of sunlight the panels will get to power the AC system and charge the batteries.

It is all a guess and the OP needs to know some days they will be cool and other days they will be hot.

Anyone expecting a perfect system to stay cool every hot day using solar panels and a battery system is having a pipe dream.

As you have stated many times, "take what you want and scrap the rest"

Pretty much my point. System design always involves specifying the duty, or what function(s) the equipment is expected to perform.

This application is no different in that sense than any other PV application: Specify the loads first. That those loads may be varying is no different than other applications.

As for a SWAG at an unknown heat gain to calc an A/C load, that's pretty easy: Put a kill-a-Watt meter on the unit and check usage on a warm day while keeping an eye on outside and inside air temps. Calc a per deg. heating or cooling load by dividing total kill-a-Watt meter read by the temp. diff. achieved, divide by the number of hours measured and multiply by 24 hrs. for a per degree day load.Then, simply size for seasonal heat spells of 3 or 4 or however many days you want. That's one way to calc a dwelling load. Works for most any dwelling. Done it many times. Grade school math.

As for how array size and batteries match up, that's a bit more involved, but the array output over, say, 4 hot days is from , say 1000 to 1400 hrs. solar time is"likely" to be higher than 4 cold days, and because hot days probably tend to be sunny either winter or summer, with, as I wrote, probably lower and less frequent winter cooling loads. Sizing under those conditions and assumptions is a bit different than using the "usual" worst solar days, because A/C loading may/may not be necessary cloudy weather. One of the uncertainties of life.

Also, as I suggested, this is not rocket science and a few simplifying assumptions will go a far piece. That's not to say however, that some reasonable use of known and simple engineering can't be put to good and easy use, particularly when estimating application loads.

[Mike90250]

> I see that solar panels are less efficient when hot, so I guessed 80w out of a 100w panel.
Almost. Generally, panels are 80% of nameplate. When really hot, only 60-70% of nameplate. Flex panels even less efficient.

[Logan5]

Not to mention 5000 BTU is a very small refrigeration unit. in a camper or RV only capable of cooling a very small area.