Inverter Sizing and Configuration for UK Solar PV Systems

Inverter Sizing and Configuration for UK Solar PV Systems: What EPC Contractors Get Wrong

Inverter sizing and configuration for UK solar PV systems is one of the most consequential design decisions an EPC contractor makes, yet it is routinely underestimated. According to Solar Energy UK, commissioned capacity on the GB grid surpassed 15 GW in 2023, yet independent post-commissioning audits continue to identify inverter misconfiguration as a leading cause of underperformance, DNO rejection, and costly re-engineering. A poorly justified DC/AC ratio, an MPPT window that clips at low irradiance, or a firmware version that fails a G99 anti-islanding test can add weeks to grid connection approval and thousands of pounds to project costs. This guide is written for experienced EPC contractors and design engineers who need a technically grounded, UK-specific workflow for inverter selection, string configuration, and simulation validation.

DC/AC Ratio Selection: Justifying the Number with PVsyst Clipping Loss Analysis

The DC/AC ratio is the relationship between the total DC nameplate capacity of the PV array and the nominal AC output power of the inverter. In standard UK practice, most EPC contractors target a ratio between 1.1 and 1.35 for ground-mount systems and slightly lower for constrained rooftop installations. But quoting a ratio without simulation evidence is no longer acceptable, particularly where a DNO technical schedule requires a formal energy yield assessment.

The practical logic behind oversizing the array relative to the inverter is straightforward: UK irradiance is relatively low on an annual basis, so the inverter operates at or near its AC rated capacity for only a small number of hours per year. Oversizing the DC side captures more energy during low-irradiance shoulder hours without meaningfully increasing clipping losses during peak generation periods.

The critical threshold most DNOs and lenders now require is that annual clipping losses do not exceed 1.5% of total energy yield. In PVsyst, this is reported in the loss diagram as “inverter overload loss” or “power limitation loss.” The acceptable threshold can vary by DNO region: a site in the Scottish Highlands with lower peak irradiance can typically justify a ratio of 1.35 or higher before clipping becomes material, while a south-facing ground-mount in the East Anglian flatlands with a DNO export limit of 50 kW per phase may require the ratio to be walked back to 1.15 to stay within both the clipping threshold and the grid connection constraint simultaneously.

The workflow in PVsyst for validating the DC/AC ratio begins in the “System” tab by entering the inverter nominal power and then adjusting module count until the simulation produces a clipping loss figure below the agreed threshold. The key mistake engineers make at this stage is failing to account for near-shading losses and soiling simultaneously, which reduces effective DC output and artificially suppresses clipping loss figures in the simulation. Running the simulation with and without shading active gives the real operational envelope. For a deeper look at how simulation tools handle shading, the PV*SOL shading analysis standard guide provides a practical comparison of approaches.

String Inverter vs Central Inverter: The Trade-Off That Defines 25-Year Asset Performance

The string inverter vs central inverter UK debate is not simply a matter of upfront cost. For any project where the asset life target is 25 years, the choice has compounding implications for MPPT granularity, fault resilience, degradation behaviour, and O&M cost trajectory.

Protection Topology and Fault Resilience

A central inverter architecture on a 500 kWp ground-mount concentrates generation risk. If the single large inverter fails or requires firmware maintenance requiring a full shutdown, the entire array is offline. A string inverter architecture, by contrast, distributes that risk across multiple units. Losing one 25 kW string inverter in a 500 kWp system takes offline roughly 5% of capacity rather than 100%. For UK projects operating under a power purchase agreement with guaranteed floor pricing, partial availability during inverter fault is a commercially meaningful difference.

Central inverters, however, offer a simpler protection topology at the AC collection point. There is a single grid interface, a single protection relay, and a single point of G99 compliance verification. For projects where the DNO technical schedule requires highly specific protection settings, one configured and type-tested central inverter can be administratively simpler to manage through the grid connection approval process than a fleet of string inverters each requiring individual firmware validation.

MPPT Channel Count and Partial Shading Behaviour

Modern three-phase string inverters marketed for commercial UK installations typically offer between two and twelve independent MPPT inputs per unit. This granularity is commercially significant on rooftop systems with multiple roof pitches or obstructions. A real-world example: a 200 kWp rooftop installation across a distribution warehouse in the East Midlands had three distinct roof orientations, including a sawtooth north-light section. Using a central inverter with a single shared DC bus, the MPPT algorithm was forced to compromise between three competing maximum power points, producing a simulated annual yield loss of approximately 4.2% compared to a string inverter configuration with individual MPPT channels per orientation. The client had originally specified a central inverter on cost grounds. After modelling the yield differential over 25 years against the O&M cost saving from a single central unit, the string inverter configuration proved more economical by year eight.

Central inverters with multiple MPPT inputs do exist and are increasingly common in the 500 kW to 1 MW range, but they remain less granular than a distributed string inverter fleet and typically require longer lead times for firmware updates that affect MPPT behaviour.

Degradation Behaviour and O&M Cost Over 25 Years

String inverters typically carry a standard warranty of five to ten years, with extended warranties available up to fifteen or twenty years. Central inverters in the same power range often carry similar warranty structures but with higher absolute repair costs due to component complexity. The O&M cost implication that most EPC contractors underestimate is the cost of a mid-life central inverter replacement. At year twelve to fifteen, a central inverter approaching end of life may require a full unit replacement, whereas string inverter fleets can be replaced one unit at a time, spreading capital expenditure and avoiding single-point failure risk. This consideration should be explicitly modelled in any financial model submitted to a project funder.

String Length Limits: How UK Module Temperatures Change the Calculation

Determining safe string length limits in a UK context requires applying a temperature correction methodology that reflects actual UK site conditions rather than the Standard Test Condition assumptions embedded in module datasheets. This is where BS EN 60904-5, the standard for low-irradiance correction, becomes directly relevant to solar PV inverter design UK practice.

Maximum Open-Circuit Voltage at UK Minimum Temperatures

The maximum string voltage calculation is governed by the module open-circuit voltage (Voc) corrected for the lowest expected ambient temperature at the installation site. In the UK, this ranges from approximately minus 10 degrees Celsius in exposed Scottish upland sites to approximately minus 3 degrees Celsius in sheltered southern English locations. Using the temperature coefficient of Voc from the module datasheet and applying the correction over this temperature range, engineers calculate the maximum Voc per module and then determine how many modules can be connected in series before exceeding the inverter’s maximum DC input voltage, typically 1,000 V or 1,500 V for commercial systems.

The error that repeatedly appears in design reviews is using the default STC temperature of 25 degrees Celsius without site-specific correction. On a Scottish site at minus 8 degrees Celsius, a typical 400 Wp monocrystalline module with a Voc temperature coefficient of minus 0.28% per degree Celsius will produce a corrected Voc approximately 9.2% higher than its STC value. Failing to account for this can push string voltage beyond the inverter’s rated maximum, triggering overvoltage protection or, in worst cases, causing inverter damage that voids the manufacturer warranty.

MPPT Voltage Window and Low-Irradiance Start-Up

At the other end of the voltage range, BS EN 60904-5 becomes relevant for low-irradiance correction. UK morning start-up conditions in winter can involve irradiance levels below 50 W/m2, at which point module operating voltage and current deviate significantly from STC values. If the string’s operating voltage at this irradiance level falls below the inverter’s MPPT minimum voltage threshold, the inverter will not begin exporting until irradiance increases. For UK sites where low-irradiance generation accounts for a meaningful proportion of annual yield, this can represent a measurable energy loss. Selecting an inverter with a low MPPT start voltage, ideally below 200 V, improves energy capture during UK winter mornings and is a differentiator that should be explicitly evaluated during the EPC contractor inverter selection process.

For a comprehensive checklist that covers module and inverter parameter collection as part of pre-design preparation, the solar design preparation checklist provides a structured framework used by design teams before entering PVsyst.

Grid Compliance Under G98 and G99: Inverter Firmware Is Not Optional

Grid compliance requirements are among the most practically demanding aspects of solar inverter configuration EPC work in the UK. The Engineering Recommendation G98 applies to installations up to 16 A per phase (typically below 3.68 kW per phase), while G99 applies to all larger installations and requires a formal application to the DNO, including a protection settings schedule and type-test evidence for all protection devices including the inverter.

Anti-Islanding Settings and Type-Test Evidence

Under G99, the inverter must be capable of detecting loss of mains (LOM) and disconnecting within specified time and frequency thresholds. The standard requires a Vector Shift or Rate of Change of Frequency (ROCOF) relay function, or an approved alternative. Most major inverter manufacturers publish type-test certificates demonstrating compliance with the Energy Networks Association’s G99 type-test schedule, but EPC contractors must verify that the specific firmware version installed on site matches the version tested. A firmware update issued by the manufacturer after type-testing that modifies protection relay parameters can technically render the unit non-compliant until a new type-test is completed.

This is not a theoretical concern. During a grid connection approval process for a 350 kWp industrial rooftop in the West Midlands, a mid-project firmware update pushed by the inverter manufacturer to address a separate cybersecurity vulnerability changed the ROCOF threshold by 0.02 Hz per second. The DNO’s technical engineer identified the discrepancy during the protection settings review, resulting in a six-week delay while the manufacturer provided updated type-test documentation. The lesson is that firmware version control must be treated as a formal project document with change control procedures, not a routine IT update.

Reactive Power Capability: Q(U) and Q(P) Curves

For larger G99 installations, particularly those connecting at 11 kV or above, DNOs increasingly require inverters to provide reactive power support using Q(U) voltage-reactive power curves or Q(P) active-reactive power curves. These settings are configured in the inverter firmware and must match the parameters specified in the DNO’s technical schedule. Not all inverter models support the full range of Q(U) gradient settings required by all DNO regions. Before specifying an inverter for a project requiring reactive power capability, the EPC contractor must confirm that the inverter firmware supports the specific Q(U) curve gradient and deadband settings specified in the DNO acceptance letter.

The process for achieving grid connection approval in the UK requires careful coordination between the inverter’s protection settings, the DNO’s technical schedule, and the project’s single-line diagram. The guide to passing grid connection approval covers this coordination workflow in detail and is essential reading before submitting a G99 application.

Inverter Configuration Workflow in PVsyst: Multi-MPPT, Clipping Validation, and IEC 62446-1 Documentation

The PV inverter sizing guide would be incomplete without a step-by-step look at how PVsyst handles multi-MPPT inverter configuration and how the output feeds into IEC 62446-1 compliant documentation.

Setting Up Multi-MPPT Inputs in PVsyst

When configuring a multi-MPPT inverter in PVsyst, each MPPT input is treated as an independent sub-array. The software allows the engineer to assign different orientations, tilt angles, and string counts to each MPPT channel, which is essential for accurately modelling rooftop systems with mixed orientations. The process begins in the “System” tab by selecting the inverter model and then using the “Multi-MPPT” configuration option to define the number of active inputs and the string assignment for each.

A common configuration error at this stage is assigning strings of different lengths to the same MPPT input. While some inverter manufacturers permit a small mismatch (typically plus or minus one module) between strings sharing an MPPT channel, PVsyst does not natively model the mismatch loss this creates. Engineers must manually calculate the mismatch loss and add it as a custom loss parameter in the “Losses” tab to maintain simulation accuracy.

Validating Clipping Loss Below 1.5% Annual Energy

After completing the system configuration, running a full annual simulation produces the loss cascade diagram. The inverter overload loss (clipping loss) is expressed as a percentage of gross annual energy. The target threshold of 1.5% is not a universal standard but has become the de facto acceptable limit for UK commercial projects under lender technical due diligence. Some lenders operating in the UK renewable debt market apply a stricter threshold of 1.0% for projects with export-limited grid connections, where clipping losses at the inverter compound with export curtailment losses to materially reduce revenue.

If the clipping loss exceeds the threshold, the options are to reduce the DC/AC ratio by removing modules, increase the inverter nominal power by selecting the next size up in the manufacturer’s range, or accept the clipping loss with a documented technical justification. The documentation of this decision should reference the PVsyst simulation report, the DNO grid connection offer, and the DC/AC ratio selection rationale as a package for the project’s technical file.

Cross-Checking Against AutoCAD Single-Line Diagrams for IEC 62446-1

IEC 62446-1 requires that the commissioning documentation for a grid-connected PV system includes a single-line diagram that accurately reflects the as-built installation. For inverter configuration, this means the single-line diagram must show each MPPT input with its associated string count, string length, and combiner arrangement. The PVsyst system configuration should be cross-checked against the AutoCAD single-line diagram before commissioning to ensure that the modelled configuration matches the physical installation.

Discrepancies between the PVsyst model and the as-built diagram are a common finding during lender technical audits. A typical example is a string being connected to MPPT input 2 on site while the PVsyst model has it assigned to MPPT input 1, which has a different orientation. This changes the modelled yield for that sub-array and can affect the bankability of the simulation report. For a comprehensive overview of how PV solar system design documentation should be structured from initial concept through to IEC 62446-1 compliance, the guide on how a PV solar system is designed provides a full design stage breakdown.

Practical Inverter Selection Criteria for UK EPC Contractors

Translating the technical analysis above into an actionable inverter selection process requires a structured evaluation framework. The following criteria should be applied during the EPC contractor inverter selection stage for any UK commercial or industrial solar PV project.

• G99 Type-Test Certification: Confirm that the inverter model and specific firmware version hold a current G99 type-test certificate from an accredited test body. Verify that the certificate covers the protection settings required by the project’s DNO technical schedule.
• MPPT Range and Low-Irradiance Start Voltage: Evaluate the MPPT voltage window against the string voltage range calculated for the site’s temperature extremes. Prioritise inverters with a low MPPT start voltage for UK winter performance.
• Reactive Power Configuration: Confirm that the inverter firmware supports the Q(U) and Q(P) curve parameters specified in the DNO acceptance letter, including gradient, deadband, and hysteresis settings.
• Maximum DC Input Current per MPPT: Cross-check the inverter’s maximum input current per MPPT channel against the short-circuit current of the planned string configuration, including the temperature and irradiance correction factors for UK conditions.
• Firmware Version Control Policy: Request the manufacturer’s firmware release policy and confirm whether firmware updates require a new G99 type-test or are covered by the existing certification envelope.
• O&M Spare Parts Availability: For 25-year asset life planning, confirm spare parts availability commitments and the manufacturer’s service centre coverage in the UK, particularly for projects in Scotland or Northern Ireland where logistics timescales can extend fault rectification periods.
• Monitoring and SCADA Integration: Confirm that the inverter’s communication protocol is compatible with the project’s monitoring platform and that per-MPPT generation data is exportable for performance ratio analysis.

DC/AC Ratio and Inverter Sizing: A UK Regional Perspective

The DC AC ratio solar UK conversation cannot be entirely divorced from regional irradiance patterns. A site in Inverness operates under a fundamentally different irradiance profile than a site in Cornwall, and the optimal DC/AC ratio reflects this. Using Meteonorm or PVGIS climate datasets within PVsyst, engineers can quantify the number of hours per year during which a given string configuration will push the inverter into clipping at a specific DC/AC ratio.

A useful rule of thumb drawn from real project data across the UK is that for south-facing, unshaded ground-mount systems, a DC/AC ratio of 1.25 is achievable with clipping losses below 1.5% for sites north of the Humber, while sites in the south of England may need to drop to 1.15 to 1.20 to maintain the same clipping loss threshold. East and west-facing arrays, due to their inherently lower peak irradiance, can often accommodate ratios of 1.35 or higher without material clipping, a configuration that can be commercially attractive when module costs are falling faster than inverter costs.

For projects with a DNO-imposed export limit, the inverter’s AC output is effectively capped regardless of DC array size. In this scenario, the DC/AC ratio calculation must account for the export limit as an additional constraint. A 500 kWp array connected to an inverter with a nominal AC output of 400 kW but constrained to a 350 kW export limit creates a layered clipping situation that must be modelled explicitly in PVsyst using the “limited power” function to produce an accurate energy yield.

Frequently Asked Questions

Conclusion: Building a Defensible Inverter Design Package

Inverter sizing and configuration for UK solar PV systems is a multi-layered engineering discipline that extends well beyond selecting a unit with the right kilowatt rating. The decisions made during inverter selection, from DC/AC ratio justification and MPPT window validation through to G99 firmware compliance and IEC 62446-1 documentation, have direct consequences for energy yield, grid connection timelines, lender due diligence outcomes, and 25-year O&M cost trajectories.

The most resilient approach for EPC contractors is to build a complete technical package for each project that links the PVsyst simulation report, the inverter type-test certificate, the DNO technical schedule, and the AutoCAD single-line diagram into a single coherent design record. Any change to one element, whether a module substitution, a firmware update, or a revised DNO export limit, should trigger a formal review of all related components.

The detail that separates a bankable design from a rejected one is rarely the choice of inverter brand. It is the rigour of the analysis behind the configuration decisions and the clarity with which those decisions are documented. For EPC contractors building that rigour into their design workflow, the technical resources linked throughout this guide provide a practical foundation for each stage of the process.

Heval Arcan