Weather-Corrected_Performance_Ratio NREL(4)

? Power meter(s) at each delivery point as defined in the contract [kW].

? (2) Calibrated reference cells or reference modules to determine POA irradiance with a target measurement uncertainty of 3%. Reference cell/module technology (poly, crystal, or thin-film) should match installed panels, but other irradiance measuring devices may be used if agreed to by all parties. Reference cells are recommended if this device will also be used to trend long term performance [W/m2].

? (1) Anemometer to measure wind speed [m/s].

? (2) Ambient temperature measurements with an accuracy of ± 1°C.

Data from the following instrumentation will be collected for reference only:

? (1) Calibrated pyranometer to measure horizontal irradiance with a target measurement uncertainty of ± 3%.

? (2) Type-T surface-mounted shielded thermocouple to measure module temperature (with a measurement uncertainty of ± 1° C) or an equivalent resistance temperature detector (RTD) device.

? Other backup meteorological measurements.

Data will be automatically collected using a combination of station and temporary loggers and instruments with a scan rate of at least one minute. Manual data sheets will be used for any non-functioning logger data channel if there will be no increase in test uncertainty.

All collected data will be averaged into 15-minute records, and each record will be used to calculate performance results and evaluate contract guarantees. Calculation methods are stipulated in this protocol.

The performance test will be deemed successful if the measured weather-corrected PR (PRcorr) is greater than the guaranteed value (within the tolerance agreed to by the parties to the test in advance of the test). Proof of Performance

Calculation Method Calculations involve the following major steps for every data record (or time interval) where the irradiance is sufficient for inverter operation:

1. Present method to calculate operating cell temperature from ambient and meteorological measurements.

2. Determine the average PV cell temperature for the solar park from the project weather file simulation.

3. Calculate the predicted PV cell temperature from measured meteorological data for each 15-minute period.

4. Use the measured irradiance to calculate the theoretical temperature-corrected DC energy.

5. Determine the measured weather-corrected PR.

6. Compare with guaranteed values.

1. Operating Cell Temperature

The following relations are used to calculate the module operating cell temperature from meteorological data. This heat transfer model is derived from the Sandia National Laboratories paper Photovoltaic Array Performance Model by King and Boyson [10]. There are other heat transfer models that can be used to calculate the operating cell temperature from meteological measurements. What is absolutely important is that the same heat transfer model is used to calculate both the:

? average irradiance-weighted cell temperature from the project weather file [°C]

? predicted operating cell temperature.

These parameters are described below.

If not separately referenced, all engineering methods presented in this report were derived from King and Boyson [10]. The calculation is done in two steps: (a) determine the module back temperature, (b) determine the internal cell operating temperature.

Based on heat transfer theory and empirical data, the PV module back temperature can be calculated with Equation (3).

(3)

Where:

Tm = module back-surface temperature [°C]

GPOA = POA irradiance from calibrated reference cells [W/m2] Ta = ambient temperature [°C]

WS = the measured wind speed corrected to a measurement height of 10 meters [m/s] a = empirical constant reflecting the increase of module temperature with sunlight b = empirical constant reflecting the effect of wind speed on the module temperature [s/m]

Tm = GPOA * {e(a+b*WS)} +Ta

[°C]

e = Euler's constant and the base for the natural logarithm.

The term within the brackets {} is an empirically determined conduction/convection heat transfer coefficient and has units of [°C m2/kW]. The empirical coefficients presented in Table 2 are recommended by King and Boyson [10].

Table 2. Empirical Convective Heat Transfer Coefficients

Module Type Glass/cell/glass Glass/cell/glass Glass/cell/polymer sheet Glass/cell/polymer sheet Polymer/thin-film/steel

Mount Open rack Close-roof mount Open rack Insulated back Open rack

a -3.47 -2.98 -3.56 -2.81 -3.58

b -0.0594 -0.0471 -0.0750 -0.0455 -0.1130

?Tcnd (°C)

3 1 3 0 3

If needed, the values for coefficients a, b, and ?Tcnd may be recomputed for different project designs if required data is available. The empirical method used to determine these coefficients are described in Incropera and DeWitt [11].

Once the module-back temperature is determined, the cell operating temperature will be calculated using Equation (4):

(4)

Where:

Tcell = predicted operating cell temperature [°C]

Tcell = Tm + (GPOA/GSTC)* ?Tcnd

[°C]

Tm = predicted module surface temperature as determined by Equation (3) [°C]

GPOA = POA irradiance, as described above [W/m2]

GSTC = reference irradiation for the correlation; constant at 1,000 [W/m2]

?Tcnd = conduction temperature drop as presented in Table 2.

2. Average PV Cell Temperature from the Project Weather File

The project weather file (based on historical data or data measured specifically for the project) is called out in the contract as the basis for the project guarantees. Also needed is the predicted POA irradiance computed from this file. This information will be used to compute the average simulated annual operating cell temperature. This is done by computing the operating cell temperature using Equations (3) and (4) for every hour of the weather data. The next step is to calculate the irradiance-weighted average cell temperature with Equation (5).

(5)

Where:

Tcell_typ_avg = average irradiance-weighted cell temperature from one year of weather data using the project weather file [°C]

Tcell_typ_j = calculated cell operating temperature for each hour [°C]

GPOA_typ_j = POA irradiance for each hour determined from the project weather file and tracker orientation [W/m2]. This irradiance is taken as zero if the sun is not up.

j = each hour of the year (8,760 hours total).

This resulting annual average cell temperature will be a constant for all further calculations. Because this averaged value is irradiance weighted, hours with high irradiance have a larger influence than hours with low irradiance. Hours with zero sun have zero impact.

Equation (5) is developed mathematically. It is this location-specific year-average cell temperature that allows correction of the PR from measured data to that predicted when the project weather file is used in a simulation to determine the guaranteed PR. It is this value that allows the weather correction to work accurately.

One proof-of-concept test is to take the simulation and calculate the PR using the traditional method, and then calculate with the weather-corrected method described in this procedure. The values will be identical.

3. Calculate the Predicted PV Cell Temperature from Measured Meteorological Data

During the test, we will need to compute the predicted cell temperature for the measured meteorological data by using Equations (3) and (4). These equations are rewritten here for clarity:

13 Tcell_typ_avg = ∑ [GPOA_typ_j * Tcell_typ_j] / ∑ [GPOA_typ_j]

(6)

(7)

Where: Tm_i = GPOAi * {e(a+b*WSi)} + Ta_i Tcell_i = Tm_i + (GPOAi/GSTC)* 3 [°C] [°C]

i = each 15-minute period of the test measurement period where measured irradiance exceeds minimum criteria. Any change in this averaging period must be agreed to in advance, since the choice of time period has a small effect on the calculated cell temperature.