How to measure the performance of your plant?

08-01-2019

The development of the photovoltaic (PV) industry has derived in new business models and scenarios for PV projects. On the one hand, due to the price competitiveness, it is key to assure the long term performance of the PV plant for guaranteeing the profitability of the project. In fact, the yield of the 25th or 30th year becomes as important as the yield of the first year, what has obvious implications in terms of degradation of the components, O&M, warranties, replacements… On the other hand, it is already a standard to manage portfolios with larger and more dispersed installations.

However, the evolution of the PV sector has not been accompanied by a parallel development of their Quality Assurance Procedures (QAP). For example, there is room for more and better quality controls in the field oriented to guarantee the long-term performance of the installations. Moreover, it is still common to evaluate the project’s performancewith a Performance Ratio (PR) value. Even contractually, the PR value determines the acceptance of the PV plant. In our opinion, this is a clearly improvable option when dealing with large and spread portfolios and when needing to analyse the project's behaviour accurately.

PV plant evaluation with PR

As in any other science discipline, PV engineering uses performance indicators to qualify the behaviour of power plants. In particular, the standard QAP includes a performance evaluation of the PV plant at two different moments. On the one hand, during the commissioning tests, at the provisional acceptance of the installation (PAC). Then, generally after two years of operation, as a part of the final acceptance of the PV plant (FAC). The most common index used in this performance tests is the Performance Ratio (PR), defined as a relation, during a certain period, between the output energy delivered by the system and the input energy received by it or, in other words, between the final yield of the installation, YF, and a reference yield, YR, obtained from the effective in-plane irradiation (Gef). More in detail, this index is defined in IEC-61724 as:


PR equation

where Et is the energy delivered to the grid during the period t, G* is the irradiance at Standard Test Conditions (STC), P* is the nominal power of the PV generator, calculated as the sum of the STC power of the modules, and Gtef is the irradiationeffectively received at the PV system.

The PR presents the advantage of being simple: it can be calculated directly without any kind of modelling, just from the data of the energy meters, the PV manufacturer datasheet and an irradiance sensor. In fact, the choice of the reference sensor for the measurement of irradiation becomes one of the few issues to face. In this respect, the standards of the sector have been evolving, depending on whether they considered:

  • The Global Horizontal Irradiation (GHI): the reference that was initially used, and corresponds to the irradiation received by a pyranometer in the horizontal plane. However, this source was soon disregarded, due to the non-linear behaviour between it and the output energy, mainly derived from the pass from the horizontal to the tilted plane.
  • The Global TiltedIrradiation (GTI): A first modification to avoid those non-linearities was to consider the global irradiance in the plane of array as a reference, as seen by a pyranometer. However, this irradiance present the disadvantage of not corresponding to the one that actually reaches the generator, since it does not take into account the angular and spectral responses of the PV modules.
  • The effective in-plane irradiation, Gef: The most advanced procedures establish as reference the irradiance received by a PV device in the plane of array. This measurement can be provided by a reference modules or cells, and represents the real inputof the system. For more information on the correct measurement of the operating conditions you can access this news.

whichever the case, for a given PV plant and site, PR tends to be constant through the years, as much as the climatic conditions tend to repeat themselves, what makes it be suitable for technical qualification when considering yearly periods (as for the FAC tests). This way, the contractual management of the PRonly requires an agreement on the guaranteed value (derived from the initial yield assessment with a safety margin agreed among the parties involved in the project), on the solar radiation measuring device and on the long term degradation effects.

PR weaknesses

Nevertheless, PR has some significant disadvantages as an evaluation index. On the one hand, it does not differentiate between avoidable and unavoidable losses in each phase, which makes difficult to preserve the chain of responsibility. For example, it does not allow distinguishing between high temperature losses (inherent of the site climatic conditions), shadowing losses (derived from the design phase), low equipment efficiency (accountable to the equipment manufacturers) or poor maintenance (accountable to the O&M contractor). Another disadvantage is its large variability in short periods of time: as it does not consider the efficiency variation with temperature and irradiance, it varies considerablyalong the year. This characteristic makes it inadequate for short evaluation periods (as the PAC tests). Figure 1 shows an example of the weekly PR evolution throughout 2010 in a PV plant in northern Spain. The main values for the whole period are: PRmean=84.2%, PRmax=93.3% and PRmin=68.8%, in which represents a ±12% excursion along the year. Even if we restrict the analysis period to a month we can find some ±7% variation. Due to this dependency on operating conditions, PR results are somehow counterintuitive. For example, the PR result of an installation with the same quality qualification will be lower if located in southern Spain than if located in northern Germany. As a result, PR cannot be used to compare the performance of different PV plants.

Weekly PR of a PV plant in northern Spain in 2010
Figure 1: Weekly PR of a PV plant in northern Spain in 2010. Up to ±12% variability can be observed along the year. Even within a month, a ±7% variation can be registered.

Therefore, as QAPs include tests during both short and long periods of time (PAC and FAC) and given that a unique reference should be established, it is not appealing to use PR as an evaluation index, but a weather independent indicator, to properly qualify the technical status of the PV plant. Otherwise, the qualification result will vary with the climatic conditions of the qualification period.

PRSTC: an overcoming proposal?

A convenient way of facing this problem is to consider a Performance Ratio corrected at STC, PRSTC, defined as the PR of a PV plant during a hypothetic period in which it had operated permanently kept at STC:


PR STC equation

where ΔETC stands for the thermal losses and ΔEG<G* for the efficiency losses due to low irradiance levels. These losses can be calculated from measured Gef and TC values and the manufacturer datasheet information. Figure 2 shows the daily evolution of PR and PRSTC of a PV plant in SouthAfrica, throughout 2017. They present, respectively, a maximum variation of ±12% and ±3.5% (with standard deviations of 3.9% and 1.0%, respectively).

Daily evolution of PR and PR STC
Figure 2: Daily evolution of PR and PRSTC at a PV installation in SouthAfrica throughout 2017. The larger PRSTC stability is clearly observable.

The outstanding benefit for the rather low added complexity of measuring the cell operating temperature is that PRSTC is neither time nor site dependent, allowing more precise qualification of the technical quality of PV installations and to detailed inter comparisons between PV plants in different climate regions.

Finally, the coherence of the QAP requires using the same performance model as in the yield assessment to establish of a correct traceability chain along the whole project. Otherwise, the assumptions underlying the initial estimations shall not be properly verified during the construction and operation phases. Then, it is advisable to establish also the initial performance expectations and the technical specifications requirements validation thresholds in terms of the PRSTC.

Conclusions

The common practice of using the PR as the index for evaluating the performance of PV plants presents several disadvantages: it does not differentiate between internal (avoidable) or external (unavoidable) losses and cannot be usedto compare plants in different locations. In order to adapt the Quality Assurance Procedures to the requirements of the current PV projects, the new technical standards (IEC-61724) consider the use of weathered-corrected indexes. Among them, QPV considers the PRSTC as particularly advisable as it allows evaluating more accurately the long-term performance of PV plants and performing portfolio analyses.