DC oversizing, specific yield, saturation losses

[Okamian]

Hi there - first time poster, new-ish user...

I'm interested to know what users' confidence level is with estimated specific yields for systems that are heavily oversized on the DC size, that have a DC:AC ratio of, say, 1.4:1 or greater.

With crystalline modules cheaper than ever, this is becoming common practice, but there can't be much empirical evidence of how effective it is. Some inverter manufacturers are publishing white papers that utilise PVsyst simulations and are showing almost-linear increases in specific yield as DC-side installation increases.

The point at which the additional saturation losses outweigh the additional production elsewhere must eventually come, but the more immediate question is at what point do the incremental increases in production become too expensive to justify, even at $0.50/W or whatever the utility-scale price-point of PV modules is these days.

Can anyone at PVsyst (Andre?) comment on what, if any, empirical research is being done to verify the model's outputs?

I'm thinking of putting together a 'small' research project that attempts to quantify at least the incremental gains vs. increased saturation losses question, and would be interested to know if there are any specific parameters that PVsyst would want to be sure were covered in order for the findings to be considered useful for use in making any adjustments to the tool.

Thanks to any and all who respond!

Ian in Ontario.

[traj123]

Very interesting question indeed and one that has been debated for a while now.

To add to what Marvin mentioned, I think that constantly operating the inverter at a higher voltage (yet within its limits), while shifting the MPPT seems to have unintended consequences that inverter manufacturers don't seem to be very clear about. I've hear a wide range of speculation ranging from DC:AC ratios not exceeding 1.12 to ones above 1.5, and some of the impacts particularly with instantaneous spikes in current when the inverter is constantly operating in the steeper part of the curve is still an open question.

On the other hand, with Marvin's second and third points, I'm a little confused, since correct MPPT tracking would imply that the energy from the modules was never realized. This would imply that there shouldn't be a temperature derate effect due to over-powering either at the inverter or the module, since the area under the IV curve is itself reduced, so the power is never produced. So, could you please clarify your points Marvin?

[André Mermoud]

For the first question, manufacturers define Pnom = "Nominal power", valid for any temperature conditions, and Pmax = "Maximum power", attainable in some conditions (not always well defined, usually an operating period).

To my understanding and after discussion with some manufacturers, this seems to correspond to a device internal temperature limitation:

 

Pnom and PMax according to temperature

 

Now for the Array temperature increase when power decreases: this is evident from the energy balance used for the evaluation of the module's temperature:

Tarray - Tamb = ( Alpha · Ginc · (1 - Effic)) / U

where Alpha = absorption coeff., Ginc = incident irradiance and Effic = module efficiency.

However with Alpha = 0.9, Ginc = 1000 W/m2, U = 29 W/m²K, the cell temperature increase is 26.4°C for effic = 15% and 31°C for effic = 0 (open circuit).

Yes the array temperature is slightly higher, but at this time the array doesn't have to produce anything !!! (or works at reduced power).