Should I add more panels?

[Aston72]

At present, in New Hampshire, I have four 250 Watt panels that only get full sun (when it shines) six hours a day.
The four panels and their micro inverters cost me about $1600.
I do not have a very good location but electric rates here are high.
My computer servers draw about 3000 watts steady.

About how many KW a month do I get with What I have?

Debating if I should I get a few more.

[azdave]

You doing a grid-assist system?

There are plenty of websites where you can input info to come up with those numbers instead of asking us to guess. Much has to do with your location and site factors that you have not shared with us and that we can't possibly know.

[J.P.M.]

Before you do anything more, buy and read "Solar Power Your Home for Dummies". You will benefit from the self education and that's a decent primer.

Anyway, too much needed information is missing in your post to guess at an answer to your question.

BTW, that should be kWh per month, or kilowatt-hours/month, not KW a month (sic).

[ImInPhxAZ]

Put your information in the calculator on pvwatts.nrel.gov, to know aprox how much you are generating. Or, look at your utility bill if your power company reports self consumption and it was legally permitted / installed.

From what it sounds you're capable of correctly installing solar at a price of $1.6 / watt, and have come no where near offsetting your usage. The question really is, why wouldn't you install more?

We would need your rates to know more about the benefits but I'm (waiting on permission) do a system that is also around $1.6/watt installed and it is cost effective with an average rate of roughly $.11 cents here in AZ. I imagine your rates are higher, given the northeast and all.

[SEJ04]

Have you guys ever found PVWatts to underestimate energy production and/or peak power production ?

[gbynum]

Have you guys ever found PVWatts to underestimate energy production and/or peak power production ?

On an annual basis, my production is 96% of what PVWatts predicted. It didn't know about my shading; I'd say it is very close given good input.

[ILFE]

... electric rates here are high.

In Southeast Asia, everywhere I've lived anyway, rates have ranged from .195c US, to .50c US, per kWh.

[SEJ04]

On an annual basis, my production is 96% of what PVWatts predicted. It didn't know about my shading; I'd say it is very close given good input.

Thanks for the feedback gbynum.

I asked as my system, on an instantaneous basis, is producing notably more power than PVWatts predicts using their hourly estimator tool. Specifically, for a given time of year, given irradiance, given temp, etc the actual power being produced under those conditions is notably greater than the PVWatts predictor. That predictor would state that I would only saturate my inverters once or twice a year (and certainly not in January), but this has already occurred a number of times. My orientation is nearly ideal as far as azimuth, panel tilt and lack of shading (a Solmetric analysis noted ~97-98% availability).

This (meaning the power value at a given moment) is of course is only a partial contributor to the actual energy production over time (so I will see as time passes how the estimates compare with my actuals), but I found it interesting (in a good way) that the power readings appears to be running positively high.

[J.P.M.]

Have you guys ever found PVWatts to underestimate energy production and/or peak power production ?

The better question is how much different has your weather been than the weather PVWatts uses to estimate long term system output. Looking at the PVWatts help screens will show that any one month's actual system output might be off as much as +/- 30 % from the model. Yearly totals can vary by probably +/- 10%.

Using PVWatts to estimate peak power production is sort of a bastardization of the models, but if you use the hourly output option and then look for a clear day's production estimate around the date you're looking for, it'll give you a 1st order SWAG of possible system output.

[J.P.M.]

Thanks for the feedback gbynum.

I asked as my system, on an instantaneous basis, is producing notably more power than PVWatts predicts using their hourly estimator tool. Specifically, for a given time of year, given irradiance, given temp, etc the actual power being produced under those conditions is notably greater than the PVWatts predictor. That predictor would state that I would only saturate my inverters once or twice a year (and certainly not in January), but this has already occurred a number of times. My orientation is nearly ideal as far as azimuth, panel tilt and lack of shading (a Solmetric analysis noted ~97-98% availability).

This (meaning the power value at a given moment) is of course is only a partial contributor to the actual energy production over time (so I will see as time passes how the estimates compare with my actuals), but I found it interesting (in a good way) that the power readings appears to be running positively high.

Your daily and long term weather will be different than what the PVWatts model uses. See the TMY manual for particulars. And/Or, inputs to the model such as orientation, module type, system loss parameter, etc., might not be representative of the actual system. In any case, PVWatts is a design tool. It is not a predictor of performance. Read the PVWatts help/info screens and the TMY manual for why.

[SEJ04]

The better question is how much different has your weather been than the weather PVWatts uses to estimate long term system output. Looking at the PVWatts help screens will show that any one month's actual system output might be off as much as +/- 30 % from the model. Yearly totals can vary by probably +/- 10%.

Using PVWatts to estimate peak power production is sort of a bastardization of the models, but if you use the hourly output option and then look for a clear day's production estimate around the date you're looking for, it'll give you a 1st order SWAG of possible system output.

JPM-

Thanks for the feedback.

My question was just a "what have you guys seen" type of question and for the SWAG you mentioned. I have a weather station installed specifically for monitoring the PV arrays (as well as a hand held device for measuring panel temperatures) , so I can see in near real time all pertinent environmentals (including irradiance) and on a clear day, if I align my steady state actual system output on the AC side with environmental points in the PVWatts table, actual output does exceed every data point I can find in the PVWatts estimations.

So yes, this is not exactly what the models are to be used for, but I am simply curious as to how accurate the models may be on a point by point basis. Of course, whats important over time is kWh. Ill see that as the year progresses. I was simply surprised that the PEAK AC output instance (for any month/temperature/irradiance) in PVWatt for my system has already exceeded a number of times in January.

[J.P.M.]

JPM-

Thanks for the feedback.

My question was just a "what have you guys seen" type of question and for the SWAG you mentioned. I have a weather station installed specifically for monitoring the PV arrays (as well as a hand held device for measuring panel temperatures) , so I can see in near real time all pertinent environmentals (including irradiance) and on a clear day, if I align my steady state actual system output on the AC side with environmental points in the PVWatts table, actual output does exceed every data point I can find in the PVWatts estimations.

So yes, this is not exactly what the models are to be used for, but I am simply curious as to how accurate the models may be on a point by point basis. Of course, whats important over time is kWh. Ill see that as the year progresses. I was simply surprised that the PEAK AC output instance (for any month/temperature/irradiance) in PVWatt for my system has already exceeded a number of times in January.

You're most welcome.

I too have a weather station for the purpose you write of. It's a Davis Pro II+ with solar sensor located ~ 4 ft. north of my array's northern edge on the N-S array centerline with the anemometer cups about 12" higher than the highest (northern) array elevation. I also measure panels temps. with an IR thermometer, but that's a longer story than I care to recount here.

How do you convert GHI to POA irradiance ?

Just like the weather 8,760 hours ago is different than now, because actual weather at any time will probably never match the weather that PVWatts uses for its modeling, you will rarely, if ever, see the same actual hourly output as PVWatts will give for the same hour, day, week, month or year.

As the PVWatts help screens will tell you, the model's output will tend to more closely match actual output as the period of measurement gets longer. Any hour's actual output will most certainly be different than what the model's output for that same hour will be. How much different is a function of how long the time period is, and some form of some probability density distribution.

Again read the PVWatts help/info screens and the TMY manual. BTW: most of the TMY irradiance data is synthetic and generated from 26 original station measurements and other weather data observations, with a lot of gaps. Little of it is from actual measurements.