• Author / Uploaded
• Hernan Arrigone

Citation preview

PVsyst guidelines

Modelling Huawei inverters in PVsyst

Huawei Technologies Deutschland GmbH

September 2018

v1.2

PVsyst guidelines

Issue and Revision Record Revision

Date

Originator

Checker

Approver

Narrative

1.0

25/05/2018

JKH/DR/P J/GS

SR

OD

First draft

1.1

20/07/2018

GRS/JKH/ DR

SR

OD

Update after addressing client comments

1.2

11/09/2018

JKH

SR

OD

Update after further client comments

Disclaimer This document has been prepared for the titled project or named part thereof and should not be relied upon or used for any other project without an independent check being carried out as to its suitability and prior written authority of RINA Consulting being obtained. RINA Consulting accepts no responsibility or liability for the consequence of this document being used for a purpose other than those for which it was commissioned. Any person using or relying on the document for such other purpose will by such use or reliance be taken to confirm his agreement to indemnify RINA Consulting for all loss or damage resulting therefrom. RINA Consulting accepts no responsibility or liability for this document to any party other than the person by whom it was commissioned. As provided for in RINA Consulting’s proposal, to the extent that this report is based on information supplied by other parties, RINA Consulting accepts no liability for any loss or damage suffered by the client, whether contractual or tortious stemming from any conclusions based on data supplied by parties other than RINA Consulting and used by RINA Consulting in preparing this report.

September 2018 │ v1.2

PVsyst guidelines

Contents 1

Introduction................................................................................................................................ 4

2

System configuration................................................................................................................. 7 2.1 Designing the string configuration ................................................................................... 7 2.2 Inputting the string configuration into PVsyst .................................................................. 8 2.3 System warnings ........................................................................................................... 11 2.4 PAN file.......................................................................................................................... 13 2.5 OND file ......................................................................................................................... 14 2.5.1 Main parameters ............................................................................................... 15 2.5.2 Efficiency curve ................................................................................................. 20 2.5.3 Additional parameters ....................................................................................... 22 2.5.4 Output parameters ............................................................................................ 23

3

Shading Scene ........................................................................................................................ 26

4

Detailed losses ........................................................................................................................ 28 4.1 Thermal parameter ........................................................................................................ 28 4.2 Ohmic losses ................................................................................................................. 28 4.3 Module quality – LID – Mismatch .................................................................................. 30 4.4 Soiling loss .................................................................................................................... 32 4.5 IAM Losses .................................................................................................................... 33 4.6 Auxiliaries ...................................................................................................................... 34 4.7 Ageing ........................................................................................................................... 35 4.8 Unavailability ................................................................................................................. 36

5

Miscellaneous tools ................................................................................................................. 38 5.1 Inverter temperature ...................................................................................................... 38 5.2 Power factor .................................................................................................................. 38 5.3 Grid power limitation...................................................................................................... 39

6

Loss calculations outside of PVsyst ........................................................................................ 41 6.1 MPPT loss ..................................................................................................................... 42 6.2 Power factor losses ....................................................................................................... 42 6.3 Grid Power Limitation .................................................................................................... 43

Appendices A.

Minimum cell temperature at the PV plant location

B.

Module Layout

C.

Glossary

D.

Frequently Asked Questions

September 2018 │ v1.2

PVsyst guidelines

1 Introduction Huawei Technologies Deutschland GmbH (‘Huawei’) has enlisted the services of RINA Consulting Ltd. (‘RINA’) to produce guidelines for replicating the characteristics of Huawei inverters within the PV yield modelling software, PVsyst. Using the datasheet, lab test information, research publications, PVsyst’s own guidance information and RINA experience, these guidelines will outline how best to convert the available information into PV yield simulation inputs within PVsyst. Different inverters require different modelling outputs: central inverters will be modelled differently from string inverters, which will be modelled differently from micro-inverters. Additionally, every inverter family and model will have different operational parameters and it is important that these are modelled correctly in PVsyst to ensure that the benefits of each inverter are correctly mirrored. RINA will introduce the reader to the method by which PVsyst models Huawei inverters, highlighting the key parameters/PVsyst inputs that describe the inverter and those which impact the inverter outputs. The guidelines are expected to be useful across Huawei’s range of inverters. To illustrate some of the principles in these guidelines, two inverter models are used as examples: the SUN2000-60KTL-M0 and the SUN2000-100KTL-H1. The following guidelines relate to PVsyst version 6.70 although the majority of the information contained within this document can be applied to previous versions. When a project is first created in PVsyst the following home screen is visible (Figure 1). Various project data need to be input/uploaded, including the meteorological conditions, details of the site, the site topology, PAN files and OND files. We have assumed the reader is familiar with these aspects of the basic set-up of a PVsyst project. Within the input parameters sections on the home screen, the sections that directly relate to inverters are ‘System’ and ‘Detailed Losses’, highlighted in Figure 1 below. The ‘System’ section will be described in Section 2 and the ‘Detailed Losses’ section will be described in Section 3. Once a project is correctly setup in PVsyst, a simulation can take between a few seconds and a few minutes to run, depending on the project size and complexity. A 100 MWp project simulation could take 5 – 10 minutes to run, for example. Once a PVsyst simulation is run, the energy yield forecast report presents a loss diagram (see the example in Figure 2). Some of these losses will be referred to within this guideline document.

September 2018 │ v1.2

4

PVsyst guidelines

Figure 1: PVsyst grid-connected project home screen

September 2018 │ v1.2

5

PVsyst guidelines

Figure 2: PVsyst report loss diagram

September 2018 │ v1.2

6

PVsyst guidelines

2 System configuration Within the ‘System’ window, the module string configuration must be entered for each sub-array of the PV plant. Prior to entering the details, the string design must be finalised.

2.1 Designing the string configuration If you are designing a plant then it is important that the number of modules per string and the number of strings per MPPT input are chosen such that the following parameters are true: 

The maximum voltage of the module string (Max voltage) does not exceed the maximum voltage permissible for the module and the inverter (whichever is smallest) using the following formula: 𝑀𝑎𝑥 𝑣𝑜𝑙𝑡𝑎𝑔𝑒 3% overload loss for ‘strong’ undersizing). If the PV modules operate well, there is likely to be some energy loss labelled ‘Inverter loss over nominal inverter power’ in the final PVsyst loss diagram. Solution: Huawei have not set a DC:AC ratio limit for system designs, but require that the maximum voltage across a module string and the maximum current per MPPT input are not exceeded. If an orange warning is raised, there is no need to rectify the design if the voltage and current limits are respected. For a red warning, the string configuration will need to be modified. Some of the warnings may arise from an incorrectly configured PAN or OND file. It is important to check that both of these are in line with their respective datasheets. Checking the PAN file is outside the remit of this document, however Section 2.4 looks at the parameters that should feed into an OND file.

2.4 PAN file The PAN file is used in PVsyst to describe the PV module parameters. Within the ‘System’ window, once the module is selected, the ‘Open’ button can be clicked to view and modify the module modelling parameters (Figure 6). Detailed discussion on PAN file modelling is outside the scope of this document but can be discussed with the module manufacturer or a technical advisor. Some key module-related parameters will be discussed elsewhere in this document, but there is one main check that must be made in the PAN file: in the ‘Basic data’ tab the user should check the parameters match those on the PV module datasheet. If parameters do not match, the user should modify the values.

September 2018 │ v1.2

13

PVsyst guidelines

Figure 6: PAN file – Basic data

2.5 OND file The OND file is used in PVsyst to describe the inverter parameters. Within the ‘System’ window, once the inverter is selected, the ‘Open’ button can be clicked to view and modify the inverter modelling parameters (Figure 7).

September 2018 │ v1.2

14

PVsyst guidelines

Figure 7: System – Opening the OND file

There are six tabs of data within the OND file: the first four will influence the yield modelling and we will go through each of these in turn. The last two tabs, ‘Sizes and operation’ and ‘Commercial’, can be ignored for the purposes of energy yield models with Huawei inverters.

2.5.1

Main parameters The Main Parameters tab for the Huawei SUN2000-60KTL-M0 default OND file is shown in Figure 8; in this tab, the input-side parameters are displayed on the left, the output-side parameters are

September 2018 │ v1.2

15

PVsyst guidelines

displayed on the right and the efficiency is displayed at the bottom. The parameters here should be updated to match information found on the datasheet. The SUN2000-60KTL-M0 datasheet is shown in Figure 9. We will explain how the datasheet parameters match with those in the OND file and whether any changes need to be made. Screenshots of the modified ‘Main parameters’ tabs for the SUN2000-60KTL-M0 and the SUN2000-100KTL-M0 will be shown at the end of this subsection. Figure 8: Default OND file – 60 KTL-M0 – main parameters

September 2018 │ v1.2

16

PVsyst guidelines

Figure 9: SUN2000-60KTL-M0 datasheet: matching datasheet parameters with PVsyst parameters Name in Pvsyst [‘Main Parameters’ tab unless specified otherwise] Max. efficiency [‘Efficiency curve’ tab]

Output side (AC grid)

Input side (DC PV field)

Euro efficiency [‘Efficiency curve’ tab] Absolute max. PV voltage Maximum current per MPPT Min. Voltage for Pnom Minimum MPP Voltage to Maximum MPP voltage Nominal MPP Voltage Number of string inputs [‘Additional parameters’ tab] Number of MPPT inputs [‘Additional parameters’ tab]

Nominal AC Power Maximum AC Power Grid Voltage Nominal AC current (only integers allowed in PVsyst) Maximum AC current (only integers allowed in PVsyst) Power factor, Cos(Phi) [‘Output parameters’ tab]

All parameters in bold are directly used for the PVsyst yield simulation. The rest of the parameters are information used for some tools, for example the Nominal MPP Voltage is used in the sizing tool. The OND file input parameters will change depending on the actual system grid voltage. The default OND file (Figure 8) is written for a grid voltage of 480 Vac. The modified OND file is shown for 400 Vac (Figure 10) for ease of explanation, as most parameters on the datasheet are given for 400 Vac. Explanations will be given on how to change OND file parameters when different grid voltages are required.

September 2018 │ v1.2

17

PVsyst guidelines

Figure 10: Modified OND file – SUN2000-60KTL-M0 – main parameters for 400 Vac

Explanation of changes Input side  

‘Minimum MPP Voltage’, ‘Maximum current per MPPT’, ‘Maximum MPP Voltage’ and ‘Absolute max. PV Voltage are taken straight from the datasheet; ‘Min. Voltage for PNom’ was calculated automatically by PVsyst using the equation: 𝑀𝑖𝑛. 𝑉𝑜𝑙𝑡𝑎𝑔𝑒 𝑓𝑜𝑟 𝑃𝑛𝑜𝑚 =

𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝑃𝑉 𝑝𝑜𝑤𝑒𝑟 𝑝𝑒𝑟 𝑀𝑃𝑃𝑇 𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝑐𝑢𝑟𝑟𝑒𝑛𝑡 𝑝𝑒𝑟 𝑀𝑃𝑃𝑇

where 𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝑃𝑉 𝑝𝑜𝑤𝑒𝑟 𝑝𝑒𝑟 𝑀𝑃𝑃𝑇 =





𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝐴𝐶 𝑃𝑜𝑤𝑒𝑟 (𝑝𝑒𝑟 𝑀𝑃𝑃𝑇) 60,000/6 = 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 0.9815

PVsyst takes the efficiency value from the curves in the next tab (‘Efficiency curve’) and is taken as the value for the nominal MPP voltage and nominal AC power specified (in this case, 600 V and 60 kVA respectively); ‘Power threshold’ is the minimum power needed for the inverter to start-up. The datasheet did not give this information, so the default value in the OND file can be used. PVsyst recommends using 0.5% of the inverter nominal power for this value if it is unknown, which is often a conservative estimate. It is rare for this threshold to result in any significant losses, i.e. the ‘Inverter loss due to power threshold’ in the final PVsyst loss diagram is expected to be 0.0%; The contractual specifications do not need to be entered for the purposes of an energy yield study.

Output side 

The Huawei inverters are tri-phased, as indicated by the ‘3W’ on the datasheets;

September 2018 │ v1.2

18

PVsyst guidelines

  





‘Frequency’ values are ticked as per datasheet specifications; ‘Grid voltage’ depends on site specifications, we are assuming 400 Vac in this case ‘Nominal AC power’ and ‘Maximum AC power’ are taken from the datasheet, which is appropriate for a grid voltage of 400 Vac: – Note that Huawei has stated that their inverters can operate at a voltage between 0.9x and 1x of their inverter rated grid voltage in order to meet site-specific grid code requirements. If the inverters do not operate at the rated voltage, their output power must be adjusted accordingly. For instance, should the inverters be required to operate at 0.9x their rated voltage (while operating at unity power factor), the maximum output power must be reduced to 0.9 times that stated on the datasheet, e.g.  for the SUN2000-60KTL-M0 inverter, the maximum power stated on the datasheet is 66 kVA so the user should enter a maximum AC power of 59.4 kVA, for a power factor of unity.  for the SUN2000-100KTL-M0 inverter, the maximum power stated on the datasheet is 105 kVA, so the user should enter a maximum AC power of 94.5 kVA for a power factor of unity. ‘Nominal AC current’ and ‘Maximum AC current’ are taken from the datasheet as appropriate for 400 V grid voltage. – For other values of grid voltage, these power values can be calculated automatically by PVsyst by ticking the boxes next to the values. The calculation that PVsyst performs for triphased systems is: 𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝐴𝐶 𝑝𝑜𝑤𝑒𝑟 𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝐴𝐶 𝑐𝑢𝑟𝑟𝑒𝑛𝑡 = 𝐺𝑟𝑖𝑑 𝑣𝑜𝑙𝑡𝑎𝑔𝑒 ∗ √3

Huawei has provided measured efficiency data at three voltages, so we can tick ‘Efficiency defined for 3 voltages’. To check the efficiency data held within the OND file, see Section 2.5.2.

A similarly modified OND file for the SUN2000-100KTL-M0 inverter is shown in Figure 11.

September 2018 │ v1.2

19

PVsyst guidelines

Figure 11: Modified OND file – SUN2000-100KTL-M0

2.5.2

Efficiency curve Within this tab, the inverter efficiencies can be defined. Inverter efficiencies can be set in two ways in PVsyst (Figure 12), with Method 2 being recommended if sufficient information is available: 



Method 1: An automatic profile can be set based on overall inverter efficiencies as found on the datasheet – this should only be used if no detailed efficiency curves are provided for the inverters; Method 2: For three DC input voltages, efficiency as a function of AC output power can be entered – this method should be used if efficiency curves at three voltage have been provided for the inverter model in question.

The steps that must be taken to correctly enter the efficiencies when using Method 2 are described below.

September 2018 │ v1.2

20

PVsyst guidelines

Figure 12: OND file – Efficiency curve

Method 1

Method 2

Huawei has provided data for the SUN2000-60KTL-M0 inverter and the SUN2000-100KTL-M0 inverters (Figure 13). For the SUN2000-100KTL-M0 inverter, efficiency curves for five DC input voltages were given. PVsyst only allows three curves to be inputted, in which case the minimum (880 Vdc), a midpoint curve (e.g. 1,080 Vdc) and the maximum (1,300 Vdc) should be used. When obtaining this data from Huawei, it is important that the data pertains to the same inverter model as installed in your PV plant. Figure 13: Efficiency data for SUN2000-60KTL-M0 and SUN2000-100KTL-H1 inverters

In order to enter data: 

‘Input voltage’ - input the three DC voltages you will enter the curves for. For the SUN200060KTL-M0 inverter, these would be: 850 V, 720 V and 600 V (as correct in Figure 12);

September 2018 │ v1.2

21

PVsyst guidelines



 

2.5.3

‘Display mode’ - As we have received data in the form of Efficiency as a function of output power, we want to display ‘Efficiency = f (P Out)’ in the Display mode box to check that the curve looks as we expect; ‘Units’ - Ensure the power units are correct for your data (kW in this case); Select either High voltage, Medium voltage or Low voltage in the ‘Input Voltage’ box before manually entering the efficiency values in the ‘Values’ box on the right. The example in Figure 12 is correct for the High voltage curve of the SUN2000-60KTL-M0 inverter. Up to eight data points can be entered per voltage level.

Additional parameters The ‘Additional parameters’ tab enables the user to enter details about the inverter’s MPPT capabilities, whether it is transformerless or not, any master/slave capabilities, information on internal power consumption and other miscellaneous electrical specifications. The important features to check are the MPPT specifications. The SUN2000-60KTL-M0 and the SUN2000-100KTL-M0 inverter both have six MPPT inputs, each able to receive data from two strings, so check this is reflected correctly as shown in Figure 14. Figure 14: OND file - Additional parameters

The other settings in this tab can be left at their default settings. Note that the small 2 W night consumption is unlikely to result in any displayed loss in the PVsyst loss tree that results once a simulation is run.

September 2018 │ v1.2

22

PVsyst guidelines

2.5.4

Output parameters The ‘Output parameters’ tab outlines the power factor capabilities and the behaviour of the inverter output at high temperatures. Figure 15:

OND file – Output parameters

Power factor The SUN2000-60KTL-M0 and the SUN2000-100KTL-M0 Huawei inverters can operate at a range of power factors (PFs): from 0.8 leading to 0.8 lagging as stated in the datasheets. These figures relate to the Cos(Phi) value between the active and reactive power. To model this feature correctly:   

Ensure that ‘Allows power factor specification’ is ticked then confirm that Cos(Phi) min and max both read 0.8, as per Figure 15; For the Huawei inverters, the nominal AC power should be defined as the active power if a power factor of between unity and 0.9 leading/lagging is to be used on the plant. If a power factor of less than 0.9 is to be used, the nominal AC power should be defined as apparent power because as soon as a reactive power requirement exists, active power will reduce but overall, the total apparent power will remain the same.

September 2018 │ v1.2

23

PVsyst guidelines

Temperature derating Regarding derating of the output power at high temperatures, some basic information relating to performance at different temperatures is provided within the inverter datasheets. For more detail, Huawei has additionally provided RINA with temperature derating data tables for both the SUN2000-60KTL-M0 and the SUN2000-100KTL-M0 inverters at different DC input voltages. The information from the datasheet and data tables are compared in Figure 16. Only one set of derating data can be entered into PVsyst, so the curve for one DC input voltage must be chosen. Input voltage will depend on design and site-specific conditions, for example string-length or temperature. Given this variation, it is prudent to consider the most conservative temperature derating curve for modelling in PVsyst. Figure 16: Temperature derating curves Inverter

Information Data Source

60KTL-M0

Datasheet

Operation Temperature Range: -25°C to 60°C

Data table from Huawei

100KTL-H1 Datasheet

Rated AC Active Power: 100,000 W @ 40°C Max. AC Active Power: 105,000 W@ 35°C Operation Temperature Range: -25°C to 60°C

Data table from Huawei

The following steps should be undertaken to model the temperature derating in PVsyst, with these settings being displayed in Figure 17:  

   

Tick ‘Allows overpower’; ‘Nom AC Power’ - enter the limiting temperature for the nominal inverter power. In the SUN2000-60KTL-M0 data table, the temperature limit for 60 kW is not listed, so a linear interpolation between the temperature limits for 65 kW and 50 kW can be calculated; ‘Max AC Power’ - enter the temperature up to which the inverter can operate at its maximum inverter power; Tick ‘High temperature limitation’; ‘Power limit abs’ – enter the power at the upper limit of the inverter’s operation temperature range. This upper limit is 60°C for both inverters; ‘Power limit #1’ – enter a data point for a temperature above that entered in the ‘Nom AC Power’ box but below the upper limit of the inverter’s operation temperature. For the SUN2000-60KTLM0 inverter, no data point has been provided between these limits therefore, again, a linear interpolation was made between the 65 kW and 50 kW data points.

September 2018 │ v1.2

24

PVsyst guidelines

Figure 17: Temperature derating of the SUN2000-60KTL-M0 inverter (left) and the SUN2000-100KTL-M0 inverter (right)

September 2018 │ v1.2

25

PVsyst guidelines

3 Shading Scene The user must model the layout of the plant so that the amount of incident irradiance on the modules can be accurately estimated. This is done by clicking on the ‘Near Shadings’ button (Figure 18) and then “Construction/Perspective” (Figure 21), which opens a window in which the shading scene can be constructed. The PVsyst website contains various sources of help for constructing a shading scene, a good place to start is here: http://files.pvsyst.com/help/near_shadings_tutorial.htm. We advise the user that for large projects especially, PV tables can be modelled as representative blocks, i.e. the exact positioning of strings horizontally relative to each other does not matter as the correct modelling of inter-row spacing, ground topology, external shading objects and division of PV tables into individual strings. Figure 18:

Near Shadings

September 2018 │ v1.2

26

PVsyst guidelines

Figure 19:

Near Shadings

Once the scene is constructed, the user must ensure that shadings are considered in the yield analysis by going to the “Use in simulation” box (Figure 19) and selecting either:  

‘Linear shadings’ if a module has cells along the length of the module (many thin-film modules), in this case row-to-row shading will affect all cells equally; ‘According to module strings’ for all other modules (including all c-Si modules on the market). A ‘Fraction for electrical effect’ of 100% implies that when a module string experiences any shade it ceases production. Presence of bypass diodes, arrangement of modules strings and distribution of shading may cause this number to be reduced. For utility-scale solar PV utilising modules with 3 bypass diodes, a value in the range 60%-80% can typically be considered, subject to the considerations above.

September 2018 │ v1.2

27

PVsyst guidelines

4 Detailed losses Within the ‘Detailed Losses’ section (Figure 20) there are eight tabs. We will describe them all in this section; however the following two tabs govern losses that are directly related to the inverter type:  

Ohmic Losses; Module quality – LID – Mismatch.

Figure 20: Detailed Losses

4.1 Thermal parameter In the ‘Thermal parameter’ tab (Figure 20, the thermal loss factors should be changed according to the installation. For most utility-scale ground-mount PV plants, ticking ‘Free’ mounted modules with air circulation will set suitable default thermal loss factors.

4.2 Ohmic losses Power is lost as it travels through the cables on the PV plant. This loss is related to the length and cross-section of the wiring. We would expect the plant to be designed in order to minimise the length of the cables and optimise the size of the cables. These losses are distributed across the cables of the plant as follows. 

DC ohmic wiring losses between the modules and the inverters. We expect variation of this loss depending on the set up of the project: – String inverter configuration: Losses in string cables from the modules to the inverter.

September 2018 │ v1.2

28

PVsyst guidelines

–



Central inverter configuration: Losses in the string cables from the modules to the DC combiner box and in the cables from the combiner box to the inverter. AC ohmic wiring losses from the inverter to the point of connection. The AC low voltage loss is the loss in the cables from the inverter to the LV/MV transformer. We expect variation of this loss depending on the set up of the project: – String inverter configuration: Losses in cables from the inverter to AC Distribution Board and from the AC Distribution Board to the LV/MV transformer. Usually they are both placed away from each other. – Central inverter configuration: Losses in the cables or busbars from the inverter to the LV/MV transformer. Central inverters are typically installed in substations or skid containers, connected to an LV/MV transformer by short run cables or a busbar connection.

Within PVsyst we can account for the ohmic losses within the ‘Ohmic losses’ tab (Figure 21). Figure 21: Detailed Losses - Ohmic Losses

When a PV plant is designed it is usual for the cable sizes to be calculated such that a specific power loss at Standard Test Conditions (STC) is not exceeded. The voltage drop within a cable dictates how much power is lost at a specific current; therefore, it is recommended that the User ascertain the designed voltage drop at STC for all cables, enabling the calculation of an overall DC ohmic loss at STC and AC circuit loss at a STC to be entered into the ‘Loss fraction at STC’ boxes shown in Figure 21. In RINA’s experience, the range of approximate loss fractions at STC anticipated for utility-scale plants using string inverters are typically: 

0.3%-1.5% for DC circuit ohmic losses, dependent on the system DC voltage (1,000V or 1,500V; the higher the voltage, the lower the losses) and cable design philosophy. – 1,000VDC systems: 0.7% to 1.5% using 4mm2 copper cables with strings of an averaged cable length (positive + negative) of 80m to 250m; – 1,500VDC systems: 0.3% to 1.3% using 4mm2 copper cables with strings of an averaged cable length (positive + negative) of 80m to 300m.

September 2018 │ v1.2

29

PVsyst guidelines



0.8%-2.5% for AC circuit ohmic losses, highly dependent on the location of the connection point to the grid and cable specifications.

The more detailed the cable calculation applied, the closer the resulting modelling losses will be to the proposed plant. Note that temperature plays an important part in the cable calculations and therefore the final calculated PVsyst inputs will need to be for Standard Test Conditions temperature (25°C). The above figures allow a range of losses to be estimated for any site at a high level as an indication. Central inverter based system designs will typically have more DC cabling and less AC cabling than a similar string inverter system, resulting in different DC and AC losses. For the central inverter configuration modelled in the Cowdown project2 our assumption for DC cabling losses was 2.5% at STC, resulting in a 1.2% final loss for this specific project. For the AC side, we considered that the low voltage connection to the transformer, via a bus bar connection, should result in a loss of around 0.1%, with the rest of the AC circuitry contributing a further 0.2% cabling loss.

Module quality – LID – Mismatch

4.3

Figure 22 shows the ‘Module quality – LID – Mismatch’ tab. The module quality and LID sections are related to PV modules only. The mismatch sections, on the other hand, relate to the string configuration, which is inverter-dependent. Figure 22: Detailed Losses – Module quality – LID - Mismatch

Module quality and LID Module quality is a figure based on the power tolerance or binning range of a PV module. For example, a 280 W module with an advertised power tolerance of -0/+3%, could be assumed to have an effective average module quality gain of +1.5%. However, if modules from this family are sold in 5 W increments, then the power bin size corresponds to 5/280 = 1.8%, therefore modules with greater than 1.8% tolerance may be sold within the next power bin size. In this case, we would

2

RINA, Huawei SUN2000 Inverter Technical Review and Comparative Study, Phase 2: Case Study Comparison, v1.4, April 2018

September 2018 │ v1.2

30

PVsyst guidelines

recommend selecting the mid-point of the smaller of the binning ranges for an effective module quality gain of 0.9%. LID is the initial degradation of a module after installation on the field, which tends to be higher than the annual module degradation rate. For crystalline silicon PV modules, this value can typically range between 0.7% and 2.0% depending on technology specifics, so should be discussed with the module manufacturer or a technical advisor What is mismatch? Mismatch is the term given to any imbalance in current or voltage across modules in a string or between strings feeding into one MPPT input. When modules are connected in series to form a string, the current output of the string will be limited to the current output of the lowest performing module. When module strings are connected in parallel, voltage mismatch can arise if there is a difference in voltages across parallel strings running into one MPPT input, which might be caused by different voltage drops along cables and slightly different voltage outputs from module strings. Mismatch in the current has been shown to have more of an impact on performance than mismatch in the voltage. The mismatch loss output in the final PVsyst yield report is for the first year of operation. It should be noted that modules will degrade at different rates over time, potentially increasing the mismatch loss. It is important to consider this variation either in an increased mismatch loss for a specific future year or as one factor feeding into an annual degradation assumption. Mismatch value for Huawei inverters RINA has undertaken a study on mismatch losses based on the current and voltage standard deviation values from module flash test data from several PV plants. Using these values and the PVsyst mismatch tool, we have identified typical average values of mismatch that we expect to find on a PV plant under different conditions, considering variations in system configuration and assuming that the modules are not current-sorted into strings. This calculated mismatch loss should be increased if strings of different power classes are feeding into one MPPT input. Inputting strings of different power classes into one MPPT input is not recommended from a design perspective. For the Huawei string inverter, as there are a maximum of two strings feeding into one MPPT input, it is easier to avoid strings of different power classes feeding into one MPPT input on a utility-scale PV plant. The calculation based upon the flash tests gives the minimum dispersion of the current and voltage parameters, as the difference in flash-tested electrical parameters is just one source of mismatch. On a plant, thermal gradients, different voltage drops through different lengths of cable, uneven soiling and partial shading could all lead to extra mismatch loss. 

For a system containing Huawei string inverters and modules with a positive power tolerance (as stated on their datasheets), a mismatch value of 0.3% is recommended. This value takes into account the flash test data calculations and plant-related sources of mismatch: – 0.3% should be entered into ‘Modules mismatch losses’ > ‘Power loss at MPP’; – 0.0% can generally remain in ‘Strings voltage mismatch’ as the value of 0.3% accounts for all the sources of mismatch listed above.

Note that for a central inverter, higher values of mismatch loss are expected, especially when there is a large spread in PV module power. As an example, RINA modelled the energy yield for a UK

September 2018 │ v1.2

31

PVsyst guidelines

PV plant, Cowdown, using firstly Huawei string inverters and secondly central inverters. The mismatch loss figure in the Cowdown project for central inverters was 0.6%3. Partial shading considerations Partial shading events can be caused by, for example, passing clouds, which is dependent on project location, external shading objects and differences in inter-row shading between parts of the array, with the overall electrical loss dependent on the number of strings per MPPT input. We note that if unshaded and shaded/soiled module strings are feeding into the same MPPT input of an inverter then the mismatch would increase. Although clouds and other transient partial shading events may also cause an increase in mismatch loss, the magnitude of the effect is considered to be within the uncertainty of the original calculations. Fixed near shading objects could cause sufficient mismatch difference to affect plants where there are hundreds of strings per MPPT (i.e. for central inverters) however for string inverters, with few inputs per MPPT, on a utility-scale plant it is unlikely that any near shading objects will materially increase plant mismatch losses. If there are likely to be numerous differences in shading on strings feeding into the same MPPT input, this shading effect can be studied using PVsyst’s module layout tool (Appendix B), with this mismatching being part of the ‘Near Shadings – Electrical Effect’ loss in the loss diagram.

4.4 Soiling loss Soiling loss can be modelled in PVsyst on an annual or monthly basis, The user should change the soiling loss based on the specifics of the installation. From our experience, annual soiling losses can typically range from 1% to 3%, although even higher soiling effects are possible if cleaning is inadequate or in situations where very high soiling rates are present. The lower soiling loss is more appropriate for a site with ample rainfall and no dust emission sources in the vicinity, the higher soiling rate would apply to a site with little rainfall, minimal cleaning schedules or high dust emission sources in the vicinity, e.g. an open mine or a desert.

3

RINA, Huawei SUN2000 Inverter Technical Review and Comparative Study, Phase 2: Case Study Comparison, v1.4, April 2018

September 2018 │ v1.2

32

PVsyst guidelines

Figure 23:

Detailed Losses - Soiling Loss

4.5 IAM Losses The IAM Losses tab is purely related to the PV module chosen and describes how reflective the module is at difference angles of incidence. This data should ideally be obtained from the module manufacturer and entered into the PAN file, if so, ‘Uses definition of the PV module’ should be ticked (Figure 24, upper). If no information can be obtained, the user should untick ‘Uses definition of the PV module’ and select ‘Ashrae parametrization’ from the dropdown list. If the PV module datasheet states that the module has an anti-reflective coating, the default b0 value of 0.050 may be slightly conservative and a lower value of 0.030-0.040 might be a better estimate (Figure 24, lower).

September 2018 │ v1.2

33

PVsyst guidelines

Figure 24:

Detailed Losses - IAM Losses

4.6 Auxiliaries The ‘Auxiliaries’ tab allows for auxiliary losses due to miscellaneous sources of PV plant energy consumption. The auxiliary losses for naturally cooled string inverters are in general much lower than that of a central inverter. Central inverters usually have a force cooling system using either forced air, or an air conditioning system. These parasitic loads vary greatly dependant on the type of cooling that is installed in the inverter, however a range between 0.2% and 0.4% is within our expectations. There is also usually lighting and electrical supplies included within the inverter cabin as well, which can increase the consumption increasing the potential variables and making exact figures hard to establish. With natural cooling and no additional sources fed from the Huawei inverters we would expect auxiliary losses to be in the region of 0.1%.

September 2018 │ v1.2

34

PVsyst guidelines

If there are transformers used for the installation then the above figures will need to be revised to ensure that transformer consumption is considered, along with any other loads that the site may need to account for, such as air conditioning units within the transformer. We would consider that a site with transformers and a small security system to have a power consumption of approximately 4 W per kW of installed capacity. For a system with approximately 0.4% of total auxiliary losses (4 W per kW), the auxiliary losses could be entered into the tab as shown in Figure 25. The variability of auxiliary losses can be high and therefore it is recommended to undertake detailed studies based on the Project design in order to derive an accurate estimation. Figure 25:

Detailed Losses - Auxiliary losses inclusive of a transformer at 0.4%

4.7 Ageing In the ‘Ageing’ tab, PV module power degradation assumptions can be set. RINA’s practice is use PVsyst to calculate the initial yield assumption and then to consider degradation over the plant lifetime as a second step. If the user wishes to consider the yield for a specific year in the future within PVsyst, they should tick ‘Uses degradation in the simulation’ (see Figure 26), enter the year for which they wish the simulation to be performed (year 10 in the example shown) and then we recommend that the annual PV module degradation rate is entered into the ‘Model’ section (0.4% per year in this example, however the value is dependent on module technology and specifics).

September 2018 │ v1.2

35

PVsyst guidelines

Figure 26:

Detailed Losses - Ageing

4.8 Unavailability The ‘Unavailability’ tab allows the user to enter the expected fraction of time the plant will be entirely non-productive (Figure 27). Plant availability is often linked to, and guaranteed within, the EPC and O&M contracts for the lifetime of the project. We would recommend using these figures for any yield or financial modelling for the plant. We note that plants with different types of inverters may experience differences in availability due to inverter downtime. For example, a much larger portion of generation lost when a central inverter fails compared to a string inverter. Furthermore, for Huawei string inverters, we note that string-level I-V testing can be undertaken via the inverter equipment, which would reduce any downtime resulting from annual I-V curve checking as part of the O&M activities. We consider that the O&M contractor should consider such factors in their guarantees or cost offerings. We consider that unavailability can be considered outside of the PVsyst model, however the PVsyst tool could be useful if periods of plant downtime due to scheduled grid maintenance, for example, are already known.

September 2018 │ v1.2

36

PVsyst guidelines

Figure 27:

Detailed Losses - Unavailability

September 2018 │ v1.2

37

PVsyst guidelines

5 Miscellaneous tools Within the ‘Miscellaneous tools’ section (Figure 28) there are four tabs, three of which have potential relevance for modelling Huawei inverters: Inverter Temperature, Power Factor, and Grid Power limitation are addressed below. Figure 28: Miscellaneous tools

5.1 Inverter temperature Huawei string inverters are always installed outdoors so the default temperature setting “external ambient temperature” can remain ticked (as per Figure 28).

5.2 Power factor The Power factor tab allows the user to state whether there are power factor requirements on the plant, at point of connection. To enter a power factor, the following steps must be undertaken:   

‘Use power factor for grid injection’ must be ticked; The power factor should be entered in the box ‘Power factor = cos(phi)’. For the example in Figure 29, a value of 0.9 power factor has been entered; Inverter PNom must be defined as ‘Apparent power’ in line with the OND file.

September 2018 │ v1.2

38

PVsyst guidelines

Figure 29: Miscellaneous tools – Power factor

Additional information on modelling power factor losses for a project are found in Section 6.2.

5.3 Grid power limitation If the PV plant has grid power limitations enforced upon it, these can be entered in the Grid power limitation tab (Figure 30). This is the easiest way of implementing an export limitation, but the power output value upon which PVsyst applies the limitation will not take into account any losses which are applied outside of the PVsyst software (see Section 6). For this reason we recommend that this loss be calculated outside of PVsyst using the process described in Section 6.3 Whether there is any grid-connection limit at all and whether such limit is applied at inverter level or at the point at which the energy is injected into the grid will depend on regulations and conditions imposed by the network operator. Figure 30:

Miscellaneous tools – Grid power limitation

‘Limit applied at the inverter level’, also known as hard clipping, is the simplest way to ensure that the grid limitation is met. Using the 600 kW limit as an example, all the inverters running at maximum power will not be able to export higher than 600 kW in the best operating conditions. If there is further electrical infrastructure e.g. a transformer and cabling as well, then an even higher loss will be seen at the grid connection points, this can be as high as 3+% for larger plants. If for example there was a detailed cable loss calculation for the AC side which returned a loss of 1%

September 2018 │ v1.2

39

PVsyst guidelines

minimum cable loss, then this could be factored into the hard clipping. In this case the hard clipping could be increased to 606 kW for the site, as it is known that 1% will be lost through the cabling. ‘Limit applied at the injection point’, usually seen in larger plants where the cost of the additional equipment, for example a power point controller, can be easily recovered. In this instance, the inverters will generate as much as possible all the time, until the power point controller has 600 kW received at the grid connection point, it will then reduce inverter outputs to ensure this limit is never exceeded. This means that all electrical losses etc. are overcome before any loss in generation is received. Should there be any underperformance on the site, for example an inverter where some the strings are shaded causing a reduction in output, then the other inverters will no longer be clipped until the 600 kW is seen at the grid connection point.

September 2018 │ v1.2

40

PVsyst guidelines

6 Loss calculations outside of PVsyst The PR of a PV plant will include some losses that cannot be calculated within the PVsyst simulation, two of which are described in this section. To calculate the overall PR based on both the losses from PVsyst those and calculated externally, the following steps must be undertaken:  

Input the individual PVsyst losses from the PVsyst loss diagram into a spreadsheet and then add the external loss calculations to the table (Columns 1 and 2 in Figure 31); As energy passes through the system - from sunlight to energy injected into the grid - each subsequent loss acts upon the energy that remains after the previous losses are taken into account, not the total initial energy (Column 3 in Figure 31). In the example below, 90.6% of input energy reaches the inverter, meaning the next loss, inverter efficiency loss, acts upon 90.6% of the initial input energy. An easy way to calculate this in a spreadsheet, is to calculate ‘1 minus the loss value’ for each of these losses (Column 3 in Figure 31). The final PR is the product of the ‘1 minus the loss value’ figures.

It is important that the losses are not simply subtracted from 100% in turn (Column 5 in Figure 31) as this will overestimate the impact of the losses. Figure 31: Post-processing losses

Note that the Loss figures shown in Figure 31 are from a specific project, the Cowdown PV plant in the UK, which uses Huawei 33KTL string inverters4. The losses shown will vary according to the PV plant components used and their specifications, as detailed throughout the rest of this document. An example of loss estimates of the Cowdown plant using a central inverter model is given in the Cowdown report.

4

RINA, Huawei SUN2000 Inverter Technical Review and Comparative Study, Phase 2: Case Study Comparison, v1.4, April 2018

September 2018 │ v1.2

41

PVsyst guidelines

6.1 MPPT loss An additional loss in energy output occurs because the inverters cannot track the maximum power point (MPP) of the strings feeding into it with 100% efficiency. As operating conditions change, the inverter must determine the maximum power available from the module strings and adjusts the operating point as required. The accuracy of this control algorithm incurs a loss. This loss is derived from:  

The ability of the inverters to track the maximum power point as tested according to standard EN50530, in other words the MPPT efficiency; The variability in weather conditions on the site, particularly the frequency of passing clouds, which would result in a more rapidly varying MPP of the strings.

This loss cannot currently be calculated in PVsyst so RINA has developed a methodology5 to estimate the MPPT loss, based on climatic information for the site region (focussing on the ratio of global to diffuse irradiance), and detailed static and dynamic MPPT inverter efficiency information provided. The static and dynamic efficiencies provided are shown in Table 1 and are considered to be very high compared to other inverters available in the market. We input the MPPT efficiency figures into our calculation tool, and evaluate the results for climatic conditions for locations around the world. Table 1: MPPT efficiency and resulting MPPT loss EU static MPPT efficiency

Dynamic MPPT efficiency*

MPPT loss

60KTL-M0

99.98% at 600 Vdc

99.87%

0.1% (to one d.p. for all locations)

100KTL-H1

99.92% at 1080 Vdc

99.82%

0.1-0.2% (to one d.p) 0.1%: to be used for areas with high GHI including Africa and the Middle East 0.2%: to be used for temperate regions including Europe

Inverter

* average from both the low and high power measurements We note that the values are strongly inverter model dependent but that for other modern inverters, RINA has seen MPPT loss figures of up to 0.9% for temperate regions and up to 0.6% for areas with high GHI.

6.2 Power factor losses Utilities often require a PV system to absorb or provide reactive power to the grid system. There are different control modes that may be enforced through the grid code, however the increased flow of reactive power results in a non-unity power factor which is controlled at the point of

5

M. Egler, S. Gordon & P. Yim, ‘Global Method for Calculating Location Specific MPP Tracking Losses Using Available Weather Statistics’, 32nd EUPVSEC, 2016

September 2018 │ v1.2

42

PVsyst guidelines

connection. The inverter’s role is to control the amount of reactive power flow, which must be provided within its capability: increased reactive current flow from the inverter results in a reduced capability to generate active power, behaving similar to a power clipping function. For this reason, the maximum active power from Huawei SUN2000 inverters reduces from the stated maximum AC power at a power factor of 1.00 (66 kW for the SUN2000-60KTL-M0 inverter and 105 kW for the SUN2000-100KTL-H1 inverter) to the inverter’s nominal power at a power factor of 0.90. Inductive and capacitive components of the PV system balance of plant (e.g. transformers and AC cables) will impact the power factor at the inverter level required, however this is not dealt with in PVsyst, and for the purpose of a yield study it is a reasonable approximation to assume that the factor required at the point of connection can be assumed the same at the inverter level. For some fixed power factor operating scenarios, it is recommended to model the balance of plant to determine the power factor settings at the inverter. In PVsyst, the power factor can be adjusted in ‘Miscellaneous tools > Power factor’ (see Section 5.2). A non-unity power factor requirement will result in the following losses:   

Losses due to reduction in active power capability of the inverter; Additional losses in efficiency of the inverter Additional losses in AC cables and transformers due to reactive current

The inverter efficiency loss could be considered as an additional loss to be added after the PVsyst simulation has been run; however the impact of this loss, which varies according to the level of reactive power requirement, would not be as significant as the above losses and could be considered non-material.

6.3 Grid Power Limitation If the PV plant has grid power limitations enforced upon it, these can be entered in the Grid power limitation tab as described in Section 5.3, however RINA recommend that this loss is calculated outside of PVsyst by applying the process described below:    

Run the PVsyst simulation; Extract PVsyst’s simulated hourly energy output values for the plant and copy them into a spreadsheet; Apply any external losses to the hourly values; Determine whether the grid power limitation is exceeded at any point and thus calculate the resultant power loss over the year due to this limit.

September 2018 │ v1.2

43

PVsyst guidelines

Appendices A. Minimum cell temperature at the PV plant location In the project home screen, click ‘Meteo database’ (orange box, Figure 32) then click ‘Meteo tables and graphs’ (red dashed box, Figure 32). Select the Tables tab (purple box, Figure 32) and then select ambient temperature, ‘Amb. Temperature’ and Hourly values before clicking ‘Table’ (blue dashed box, Figure 32). Figure 32: ‘Meteo database’ – meteorological tables

The hourly temperature values over a typical meteorological year will pop up in a separate window. Find the maximum and minimum temperatures recorded in this list. It may help to export the file into a spreadsheet by clicking ‘Export’ then ‘Copy to Clipboard as text’. Now these values can be pasted into a spreadsheet for easier examination, and the maximum and minimum ambient temperatures can be identified. Whilst this provides an indication of the temperature profile for a typical year, extreme values may not be included here. It is therefore recommended that, should

September 2018 │ v1.2

PVsyst guidelines

local temperature information be available (national meteorological organisations, SYNOPTIC network ground station data, climatic data etc) temperature extremes are investigated in more detail to confirm any assumptions used. Next, go back to the Home screen and open ‘Project settings’ and in ‘Design conditions’ change the ‘Lower temperature for Absolute Voltage limit’ to the minimum temperature for the site, rounded down to the nearest integer. Figure 33: ‘Project settings’ – changing the temperature limits

B. Module Layout This tool enables the user to set modules into strings according to the Single Line Diagram (SLD). It can be useful to evaluate partial shading effects on mismatch calculations but should not be used for very large plants as it is computationally intensive and thus a very slow tool. Near shadings should be used to model shading for the whole field in order to determine shading losses but the module layout tool can help us identify how differences in partial shading across strings that feed into one MPPT input would affect the mismatch value. To carry out this evaluation, firstly the near shading scene needs to be modelled accurately, as near shading objects and PV table specifications cannot be changed within this Module Layout section. Once that is done, the following steps should be undertaken in the Module Layout tool:

September 2018 │ v1.2

PVsyst guidelines

‘Mechanical’ tab 

In the ‘All subfields’ tab – check the module details are correct then click “set modules” and the formerly white sheds will fill blue. Each individual module needs to be positioned correctly before other options become available so it is important that the shading scene correctly matches the dimensions of all the modules.

Figure 34: Module Layout - Mechanical

‘Electrical’ tab Within the ‘Electrical’ tab, the modules can be positioned into strings. This can be done automatically, using ‘Auto attribution’ or modules can be placed individually according to the stringing design.

September 2018 │ v1.2

PVsyst guidelines

Figure 35: Module Layout - Electrical

‘Shadings 3D’ tab Within the ‘Shadings 3D’ tab, the shading scenario can be run for one day, chosen to be mid-winter in order to identify the maximum impact on shading. Once run, ‘Shading factor for beam, electrical’ will give the ‘Near shadings – electrical effect loss’ as applicable to the string layout. This value can be compared to the ‘Near shadings – electrical effect loss’ as calculated for that day in the ‘Near Shadings’ tool, in order to calculate whether any extra loss should be considered.

September 2018 │ v1.2

PVsyst guidelines

Figure 36:

Module Layout – Shadings 3D

C. Glossary Table 2: Glossary Term

Description

Impp

Current at the maximum power point (of a PV module or a string)

Isc

Short circuit current (of PV module or a string)

I-V curve

Current-voltage curve

MPPT

Maximum power point tracker

O&M

Operations & Maintenance

OND file

Onduleur file = the file detailing the inverter parameters in PVsyst

PAN file

Panneau file = the file detailing the PV module parameters in PVsyst

PF

Power Factor

PR

Performance Ratio

String

PV modules connected in series

SLD

Single Line Diagram

STC

Standard Test Conditions

Vmpp

Voltage at the maximum power point (of a PV module or a string)

Voc

Open circuit voltage (of a PV module or a string)

September 2018 │ v1.2

PVsyst guidelines

Term

Description

Impp

Current at the maximum power point (of a PV module or a string)

Isc

Short circuit current (of PV module or a string)

D. Frequently Asked Questions The following questions were proposed by Huawei as being commonly asked questions from their customers. 1. Why do you use power sharing function since Huawei string inverter doesn’t have power sharing ability? The manuals state that the PV strings connecting to the same MPPT circuit should contain the same number of identical PV modules. The possibility of having an unbalanced input configuration is not stated. ‘Power sharing’, enables the user to adjust the percentage of nominal power per MPPT for each subarray (see Section 2.2). We refer the user to the PVsyst online help guidelines (http://files.pvsyst.com/help/powersharing.htm) for more details information if an unbalanced configuration must be modelled. 2. What’s the mismatch loss by using string solution? Although string inverter solutions offer an advantage over central inverters with respect to mismatch, the overall power loss is dependent on several factors, mainly driven by different climatic or design conditions as well as non-homogenous electrical properties of the PV array. These factors include manufacturing tolerances, thermal gradients across modules, uneven soiling, uneven module degradation, mismatched modules on a string, mismatched strings into an MPPT input and partial shading. Based on the module power tolerance i.e. if the module tolerance is positive, symmetrical or asymmetrical, our expectations for this loss in string solutions will range between 0.3% and 0.6%, in comparison, central inverter solutions are expected to have a higher mismatch loss. 3. Why do you set the mismatch value for a plant with string inverters lower than that for a plant with central inverters? How will the mismatch value change with bifacial modules? There are significantly more strings on a central inverter with each string introducing its own slightly different electrical characteristics (both from production and also installation e.g. soiling, shadings, etc.), which is compounded with the higher variance. This means that the MPP tracker will have a larger number and range of maximum power points to try and track at the same time, causing a greater loss. Bifacial modules absorb light from both sides of the module. This causes a greater variability of the amount of light absorbed by each module and thus a higher value for mismatch should be used for a bifacial PV plant compared to a monofacial PV plant. The extent of this increase in mismatch value will depend on factors such as number of PV modules mounted vertically on a table, module tilt angle and height of modules from the ground and rear shading objects.

September 2018 │ v1.2

PVsyst guidelines

4. Why the overload loss in the “system” is bigger than the inverter power loss in the “loss diagram” of report? The loss diagram figures may be significantly different from the ‘system’ sizing value because the histogram for the sizing tool is a high-level analysis that does not account for all the system losses (e.g. soiling, DC cable loss, etc). Once these are discreetly calculated in the hourly simulation, the inverter DC input is lower, which then results in a lower final inverter undersizing loss. 5. Why does the maximum current in the OND file not match the maximum current in the datasheet? Some values in PVsyst are only displayed as integers and the maximum current is one such value, when inverter power ratings exceed ~80 kVA. For example, the datasheet of the SUN2000100KTL-H1 inverter states a maximum current of 80.2A, whereas in the OND file, 80A is displayed and PVsyst does not permit ’80.2’ to be entered. This does not mean that 80.2A is not used, however. The OND file can be opened as a text file (the OND file located in the user’s personal PVsyst database, within Compos PV > Inverters) and here the user can check and even modify the maximum current parameter (see the red box in Figure 37). Figure 37:

OND file text file - modification outside of PVsyst

September 2018 │ v1.2

PVsyst guidelines

6. How do you set the configuration to distinguish the different temperature design. For example, for 40℃ or 35℃, how to set the parameter? In the ‘Main parameters’ tab of the OND file, it is possible to define different ‘Nominal AC Power’ and ‘Maximum AC Power’ values. Once these are different, it is possible to define the temperature intervals in the ‘Output parameters’ tab. In order to input correct figures, we recommend that the temperature power derating profile of the inverters are requested from the manufacturer. Huawei has provided these details. In the ‘Output parameters’ tab, the user should allow for overpower and high temperature limitations and input the relevant temperatures for each AC power value. Please see Section 2.5.4 for more details. 7. When overloading more than 40%, the “inverter loss over normal power” of string inverter is higher than central, why? Should the user define the same number of modules per string, same inverter MPPT windows and same DC/AC ratios for string and central inverters with similar efficiencies and oversizing properties, we do not foresee major deviations, as long as the total number of strings are identical. If there is a difference in the ‘inverter loss over nominal power’, the difference is likely to be due to the differences in the inverter efficiencies or oversizing capabilities. 8. How to set availability in PVsyst, what's the difference between string and central solution? Plant availability is determined by the quality of the PV components (modules, cables, inverters etc.) but also dependent upon features of the O&M contract, i.e. how well the plant is monitored and how quickly any component faults are fixed. As such, we consider that unavailability is a loss that can be applied after the PVsyst yield is run and should be based upon the specification of the O&M contract. Typically this would be considered outside of the energy yield simulation as a separate line item in the financial model. Saying that, there is an availability function in PVsyst which also allows the periods of unavailability to be defined. The effective energy loss is dependent on the season and weather conditions during the unavailability periods. Therefore, the unavailability loss has only a statistical meaning. Typical financial model assumptions for plant availability do not typically differentiate between string and central inverter technologies but there will be different factors contributing to the availability of the two types of system. A central inverter failure generally results in a higher proportion of the plant being unproductive and the failures can be more complex to repair and may require specialist labour, which may be reflected in the O&M contract price. The increased complexity and generation impact from a single unit outage can lead to central inverters contributing towards a lower overall plant availability. Conversely string inverters have an increased component count and therefore more frequent failures may be expected on-site; typically string inverter issues can be resolved by the O&M team by replacing the string inverter with spares stock, assuming number of failures are within expectations and spares are available. For Huawei string inverters, we note that string-level I-V testing can be undertaken via the inverter equipment, which would reduce any downtime resulting from annual I-V curve checking as part of the O&M activities. For a central inverter system, string-level I-V curve checking would require the string’s combiner box to be disconnected.

September 2018 │ v1.2

PVsyst guidelines

9. What is the difference in setting “Multi-MPPT feature” and inverter quantity, in system configuration? If the inverter has multiple MPPTs, you can connect strings with different characteristics to separate MPPTs to avoid further mismatch losses; for example if a PV system has different module orientations, you can connect one module orientation to one MPPT and other orientation to another MPPT. When the ‘multiple MPPT’ feature is set in PVsyst, the software models each MPPT as an independent inverter, enabling more complex system configurations to be input and more accurate inverter losses to be output. 10. Why is it best to calculate grid power limitation outside of PVsyst? Section 5.3 describes how a grid power limit can be input into PVsyst and summarises the effect of different ways of accounting for the grid power limits. Although PVsyst’s grid power limitation tool is an easy-to-use way of implementing a grid power limitation, for it to be accurate all potential sources of energy loss must have been modelled within PVsyst. It is important that grid power limits are applied to the final energy output of the PV array (or inverter, depending on where the limit is to be applied) after all losses have been accounted for. Therefore if any external losses are to be applied to the PVsyst output after the simulation has been run, the grid power limitation loss should only be calculated once these losses have been considered, which means calculating the loss separately, outside of PVsyst, as described in Section 436.3. 11. What are the P50 and P95 values in Pvsyst? How can one use these parameters? In ‘Miscellaneous tools’ there is a tab called ‘P50 – P90 estimation’, which displays a probability distribution graph once the PVsyst yield simulation has been run (Figure 38) Figure 38:



Miscellaneous tools – P50 – P90

In any single year, there is an estimated 50% probability that the energy produced by the plant will exceed the P50 energy value.

September 2018 │ v1.2

PVsyst guidelines

 

In any single year, there is an estimated 90% probability that the energy production will exceed the P90 energy value. Any Pxx value can be calculated by inputting the xx value in the ‘Resulting estimation’ box

P50 and P90 values are often requested for use in the financial models of any party investing in the PV plant. In order to obtain the P90 value, uncertainties are applied to the yield calculation, for example measurement uncertainties in irradiation or module flash testing. Values for variability and uncertainty in irradiation data and yield modelling parameters can be input within the ‘P50 – P90 estimation tool’. A detailed uncertainty analysis is likely to involve more parameters than those listed here, for example, RINA performs an uncertainty analysis outside of PVsyst to encompass other parameters such as, uncertainties due to the range of irradiation sources used, inter-annual variability in irradiation and uncertainties associated with each individual energy loss. We recommend that the user undertakes (or commissions) a detailed uncertainty analysis before considering the P90 values. Additionally, it should be noted that PXX factors are not intended to capture the potential for project underperformance, only uncertainty in the various modelling inputs and methodologies.

September 2018 │ v1.2

PVsyst guidelines

RINA Consulting Ltd. 2nd Floor Offices Nile House, Nile Street Brighton, BN1 1HW, UK +44 (0)1273 819 429 rina.org

September 2018 │ v1.2