insolation & performance ratio calculation

[davemac_314]

Hello,

I have an Apogee SP-215 silicon pyranometer and a revenue grade meter installed on a
thin film system. I've been acquiring data from the pyranomter (W/m^2) and meter (kWh)
every 5 minutes.

On the performance ratio calculation, I'm using kWh/(system size in kW) / (insolation/1000).

Currently, I'm converting the irradiance data to insolation (Wh/m^2) by getting the summation
of the irradiance data set then multiply by 1/12 i.e. converting W/m^2 to Wh/m^2 (5mins = 1/12 hr)
Is this approach accurate enough for the performance calculation?

Is it also good to calculate daily performance ratio?

Any suggestions for improvement?
Thanks,

Dave

[J.P.M.]

Hello,

I have an Apogee SP-215 silicon pyranometer and a revenue grade meter installed on a
thin film system. I've been acquiring data from the pyranomter (W/m^2) and meter (kWh)
every 5 minutes.

On the performance ratio calculation, I'm using kWh/(system size in kW) / (insolation/1000).

Currently, I'm converting the irradiance data to insolation (Wh/m^2) by getting the summation
of the irradiance data set then multiply by 1/12 i.e. converting W/m^2 to Wh/m^2 (5mins = 1/12 hr)
Is this approach accurate enough for the performance calculation?

Is it also good to calculate daily performance ratio?

Any suggestions for improvement?
Thanks,

Dave

Is the instrument in the horizontal position or in the plane and azimuth of the collector ?

Usually, and for a lot of reasons that really do make sense and that took me several years to appreciate, the instrument is placed/used in the horizontal plane and irradiance in the plane of the collector is then derived from any number of methods/algorithms available. That can get as complicated as you want and may be necessary if you want to realize the full power of the instrumentation you're dealing with.

Clear sky irradiance in the horizontal plane, as a 1st approx., is usually somewhere around 900-1000 W/m^2 as f(a lot of stuff). Depending on what your output shows, mV signal , W/m^2, Langleys, etc.,the horiz. irr. as raw data will need to be converted to irradiance in the plane of the collector. A lot of that is a matter of some interpretation and the method(s) used. If you haven't found out yet, this is not an exact science.

Suggest getting ahold of a copy of Duffie & Beckman for a primer on irradiance availability and measurement - a bit dated but the basics have not changed much over the last 150 yrs. or so since a guy named Smithson got into it.

[davemac_314]

Is the instrument in the horizontal position or in the plane and azimuth of the collector ?

Usually, and for a lot of reasons that really do make sense and that took me several years to appreciate, the instrument is placed/used in the horizontal plane and irradiance in the plane of the collector is then derived from any number of methods/algorithms available. That can get as complicated as you want and may be necessary if you want to realize the full power of the instrumentation you're dealing with.

Clear sky irradiance in the horizontal plane, as a 1st approx., is usually somewhere around 900-1000 W/m^2 as f(a lot of stuff). Depending on what your output shows, mV signal , W/m^2, Langleys, etc.,the horiz. irr. as raw data will need to be converted to irradiance in the plane of the collector. A lot of that is a matter of some interpretation and the method(s) used. If you haven't found out yet, this is not an exact science.

Suggest getting ahold of a copy of Duffie & Beckman for a primer on irradiance availability and measurement - a bit dated but the basics have not changed much over the last 150 yrs. or so since a guy named Smithson got into it.

Actually, I have three pyranometers installed. Two of them are in the plane of array, which is 8 degrees facing South. One is on the horizontal plane.
I'm using the the plane-of-array pyranometer for calculating the insolation so I think I won't get into the derivation since I already have data for it.

Is the 5min resolution acceptable based on your experience? Or would it introduce a slight error?
How about calculating performance ratio on a daily basis?
Thanks

[J.P.M.]

Actually, I have three pyranometers installed. Two of them are in the plane of array, which is 8 degrees facing South. One is on the horizontal plane.
I'm using the the plane-of-array pyranometer for calculating the insolation so I think I won't get into the derivation since I already have data for it.

Is the 5min resolution acceptable based on your experience? Or would it introduce a slight error?
How about calculating performance ratio on a daily basis?
Thanks

Under constant (cloudless) conditions, I'd guess that 5 min. intervals would not introduce a whole lot of additional error. How much is always the question. More sky variation than that and you're got more uncertainty. How much ?

I'm assuming you have 2 instruments in the plane of array to measure direct and diffuse, and all 3 are of the same type ? A respectful comment or 2:

1st, Si instruments of the type you described above may not do the best job of accurately measuring diffuse rad. intensity when used with a shadow band or shading disk because of the spectral response of Si instruments which is usually more mismatched with diffuse than beam radiation, diffuse often being of a different wavelength distribution than global beam + diffuse wavelength distribution which the device and it's software thinks its seeing.

2d, Again assuming 1 tilted instrument is for diffuse measurement - tilting Si instruments off horiz. can induce error due to the increase/change in the reflected and horizon brightening portions of the total irradiance with their different wavelength dist. because of the spectral response of the silicon cell device. Used to be w/ the Eppley instruments, which are true pyronometers (and mucho bucks which is why the silicon devices are much more common), off horizontal screwed up the temp. dist. in the domes and screwed up the output by a few + % depending on temp./tilt. The SI devices are better in that respect, but the spectral dependence must be considered since most reflected and horizon brightening terms are usually more diffuse in nature, thus a different wavelength dist. (usually shorter), adding error to the diffuse measurement. These are a couple of the reasons I was alluding to above. Add in cosine response changes and things begin to get interesting. Si devices can be used in any orientation as all the product blurbs state, but the spectral response will be different for reasons different than the old thermopiles. Some even specify +/- 5 % (or so) accuracy w/ regular recalibration, but add the horizontal requirement for that accuracy. Pay your money, take your choice. All the serious research I've ever seen used instruments in the horizontal and redefined irradiance in the collector planes in standard ways and used the instrument in the intended way.

I've got a Davis instrument and the solar sensor is Li-Cor. The logger records at 1 min. intervals. The array monitor reports at 5 min. intervals and I wish it was 1 min. like the Davis/Li-Cor. Interpolating the array 5 min. interval data down to 1 min. is adequate, but less than ideal. Probably acceptable under cloudless conditions and clear skies w/ direct beam radiation > 600W/m^2. The solution for me is to take instantaneous measurements and integrate them over the period in question. That works best under clear skies, less so for les clear skies with respect to daily totals.

[davemac_314]

HI JPM,

Thanks for the reply.
By the way, does performance ratio reach over 90% for thin film installations?
I got 90-95%. Is this acceptable for thin film?

[J.P.M.]

HI JPM,

Thanks for the reply.
By the way, does performance ratio reach over 90% for thin film installations?
I got 90-95%. Is this acceptable for thin film?

You're welcome. Hope it provided food for thought/action.

I'm not sure I understand the question. I do not understand what you mean by performance ratio. I.E., ratio of what ?

Regards,

[inetdog]

On the performance ratio calculation, I'm using kWh/(system size in kW) / (insolation/1000).

I think this is what he means, where insolation is total insolation (energy) per 24 hour period, not the incident power level.....