PV Performance Guarantees (Part 2): : Page 2 of 4

Proof of Performance & Guarantee Structures

Power over time equals energy. This basic concept is frequently misunderstood or misapplied. While a PV plant can easily be characterized by nameplate capacity alone (MWDC-STC), this is just the power part of the equation. The problem with this metric is that PV plants operate far from nameplate values most of the time, and each component of the system behaves decidedly differently under changing conditions. Energy (MWh) is ultimately what produces revenue. Understanding PV performance characteristics is key to understanding how nameplate ratings change over time.

Performance is climate-dependent. Quantitative PV performance depends on climate conditions. The two most directly indicative factors are module cell temperature and plane-of-array (POA) irradiance, but there are dozens of second-order factors. As a result, the only way to objectively evaluate PV power plant performance is to take the weather out of the analysis. This takes some very accurate measurements and relatively tricky math, as described in Part One of this article.

From these two simple concepts, two things are clear about evaluating PV plant performance. First, you must know what should happen, which is done through simulation and modeling. Second, you must know what did happen, which is accomplished by plant metrics.

PLANT METRICS

Regardless of the method being used to prove performance, all interested parties must understand the methods, expectations and protocols for plant metrics. These measurements are ultimately key to proving plant performance.

The project team must first establish a basis of design for the pieces of the system between the modules and the point of interconnection. This can take any number of forms, but within the scope of a performance guarantee, it should represent the mutually agreed upon electrical behavior and climate response of the system. The design basis should include parameters for system losses, efficiencies and valid assumptions for temperature response. Defining the basis of design is instrumental in setting the expectations for system performance.

Definitive plant parameters to include when you are drafting a design basis for a performance guarantee include dc wiring losses, transformer losses and module soiling. Rather than rely solely on inferred values for these quantities, it is good practice to compare actual values to calculated values. Partload wiring and transformer losses, for example, are drastically different than full-load values. In order to accurately consider this effect when detailing expected production values, it is very useful to use measured values that are taken once the system is operational. Module soiling, similarly, varies over time and in a decidedly nonlinear fashion, so the effects of soiling should be quantified in the field in a before-and-after approach with strategic module cleaning.

After the basis of design is established, PV plant output is primarily a function of the following measured or measurable quantities: plane-of-array irradiance, cell or module temperature, module degradation, inverter efficiency and ac collection losses.

Plane-of-array irradiance. POA irradiance is a critical metric for performance calculations because current and power are directly proportional to irradiance in the plane of the array. The sensors are placed in precisely the equivalent installed condition as the array. Undoubtedly, a high degree of measurement overlap and redundancy is needed to mitigate the effects of clouds and shading, and to check for sensor correlation. There should be a minimum of two pyranometers for every project greater than 1 MW; there should also be at least one irradiance meter for every additional MW. POA sensors are electrically similar to PV modules, so they tend to respond to irradiance much like a module does. Other methods to arrive at plane-of-array values use transposition factors and calculations to approximate POA irradiance from, for example, global horizontal values, but POA measurement instruments eliminate the need for any intermediary steps.

Module or cell temperature. Highly accurate temperature sensing is an extremely cost-effective way to help predict system performance. High degrees of sensor overlap and redundancy are easy to implement and give very good indications of operating conditions. For systems larger than 3 MW, there should be one sensor for every 500 kW of array capacity. In large arrays, the module temperature can vary greatly, however, so strategic placement of sensors is imperative to ensure that measurements accurately match average conditions. Using the information correctly is tricky and requires sophisticated solar expertise.

Article Discussion

Related Articles