Weather-Corrected_Performance_Ratio NREL(3)

To further demonstrate the variability of the PR, Figure 3 shows the hourly values of corrected and uncorrected PR for the entire year (from the same simulation used in Figure 2). Note that the extreme variance in the uncorrected PR values is drastically reduced with the weather correction.

5

Figure 3. Corrected and uncorrected PR calculated for each hour of the year for the same simulation described in Figure 2.

Average Annual Cell Temperature The important parameter for the corrected PR is the annual average cell temperature. In Equation (2) this is the variable “Tcell_typ_avg.” This value is carefully determined such that the annual PR and annual weather-corrected PR are the same for this location. While the PR shows systematic seasonal variations as shown in Figure 2 and Figure 3, the weather-corrected PR provides a consistent assessment of the annual PR with very little seasonal variation if the “Tcell_typ_avg” is chosen correctly.

The average annual cell temperature should be determined from the project weather file and the simulated plane-of-array irradiance that is used to set the expected power generation. The project weather file represents the nominal annual meteorological conditions hour by hour. As noted above, when the “Tcell_typ_avg” is chosen correctly, the annual PR is the same number as the annual weather-corrected PR. The “Tcell_typ_avg” is chosen by applying Equations (1) and (2) until the annual PR and annual corrected PR are equal, or (equivalently) using the equations provided in

this report.

6

Calculated Cell Temperature It is important that the computation method used to determine the average annual cell temperature from the project weather file is also used to compute the operating cell temperature during the PR test. It is very important to follow this requirement because it mathematically assures a link between the PR calculated from measurements and the PR predicted by the project weather file and the plant design parameters (loss factors). This linkage is broken if the cell temperature is measured directly or determined by a different method.

It should also be noted as an aside that directly measuring the operating cell temperature may hide design or construction issues that result in higher than expected operating temperatures — and thus a false pass of a performance test using the weather-corrected PR.

Therefore, the solution is to compute cell temperature from the weather data. If the computation is consistent for (a) the annual average cell temperature and (b) the current operating cell temperature, then this approach will result in consistent values throughout the year. As stated, the result is also tied back mathematically to the value calculated using the project weather data. With this method, we have a consistent basis for the weather-corrected PR. The corrected PR calculated from the field measurements is consistent with the performance guarantee.

Precise methods based on a Sandia [10] heat transfer model are used in this report for calculating PV operating cell temperature. This method was previously validated [10], so is not validated here, but it is noted that the coefficients used for this calculation may need to be chosen to match the operation expected for the system being measured. This method may be replaced with a different heat transfer model as long as identical methods are used to compute both the annual average cell temperature (from the historical weather data) and the cell operating temperature (from the measured weather data during the assessment period).

Example Data Thus far, this report contains a theoretical presentation with some simulated data examples. A logical question to ask is: does this method actually work? Real-world results are presented. Corrected Performance Ratio from Measured Data The plots shown in Figure 4 show the uncorrected (raw) and weather-corrected PR from a Provisional Acceptance Certificate (PAC) used to confirm engineering, procurement, and construction (EPC) project guarantees from around the world. Notice how the weather-corrected PR provides stability to this volatile metric.

7

4,700-kWp Facility

2,300-kWp Facility 1,300-kWp Facility

Figure 4. Weather-corrected PR compared with uncorrected PR during acceptance testing. Figure 5 shows the daily PR for a 24-MW facility over a year. This is for day-to-day operation. There is increased scatter because this data was not collected during a controlled performance test. The weather-corrected PR significantly removes the seasonal bias. Actual operating issues such as soiling or derated equipment are more apparent when considering the corrected PR.

8

Figure 5. PR (weather-corrected and uncorrected), annual trend, 24-MW facility.

Corrected PR Test Protocol The remainder of this report presents a sample protocol for a weather-corrected PR measurement. It is recommended that practitioners fully adopt the mathematical approach to ensure fidelity in the final calculated PR. However, commercial terms can be adjusted as needed for items such as number of sensors, minimum irradiance criteria, and treatment of measurement uncertainty.

Purpose The procedures below describe the calculation methodology used to determine the weather-corrected PR for a plant acceptance test. The guiding principle is that the measurement and computational approach provide a method that results in an accurate and consistent metric for determining whether performance guarantees have been demonstrated. This metric must be unbiased to test boundary conditions, and thus fair to all parties. The purpose is to measure against an annual PR guarantee.

Note that the PR should not be used as a guarantee metric if the plant is designed such that the inverter will clip during high irradiance. The PR will unfairly penalize results during these periods. Similarly, grid unavailability or other circumstances may affect the fair application of the metric, and the metric is not designed to cover reactive power requirements.

Parties to the Test and Responsibilities Parties to the test are defined in the contract. They may be the owner, contractor, and an independent engineer. All parties must agree to this protocol before test commencement. The test will be executed by the contractor. All relevant raw test data, spreadsheets, and computations shall be provided to all other parties to the test for their review. The contractor will supply raw data before any manipulation and highlight any gaps in the data. The final test report will be produced by the contractor in the timeline detailed in the contract.

9

During the test, any anomalies to this protocol will be documented. Resolutions to anomalies or variations to this protocol that occur during the test period will be documented and approved by all parties to the test in order to continue with the testing effort.

Requirements Before the Test Before the test can commence, the following need to be completed:

1. Install and calibrate all primary measurement instruments.

2. A test procedure has been published and agreed to by all parties to the test.

Minimum Irradiance Criteria The plant acceptance test period is five days long with the following minimum irradiance criteria:

? At least three days must have irradiance measured in the plane of the array that is greater than 600 W/m2 for three continuous hours, and the daily total irradiance must exceed 3,000 Wh/m2/day.

? If there are not five days that meet these minimum irradiance criteria, the test period may be extended until five sufficient days have been recorded. There will not be any liquidated damages triggered as a result of this weather-related test delay.

? If there are not five days that meet the minimum irradiance criteria, yet the corrected PR of the five strongest days meet the contract guarantee, then the plant acceptance test will be deemed a success.

Instrumentation Data from the following instrumentation will be used to determine park performance:

bbs.99jianzhu.com内容：建筑图纸、PDF/word 流程，表格，案例，最新,施工方案、工程书籍、建筑论文、合同表格、标准规范、CAD图纸等内容。