Solar PV and Battery Storage in HEM: Technical Guide

The Home Energy Model (HEM) models solar PV generation and battery storage at half-hourly resolution, a transformative improvement over SAP's monthly approach. Using the methodology defined in HEM-TP-18, HEM calculates PV output at each of 17,520 timesteps per year based on actual solar irradiance data, panel orientation, tilt angle, shading, and inverter efficiency. It then determines how much of that generation is consumed on site, how much is stored in batteries, and how much is exported to the grid. This is the first time a UK regulatory energy model has been able to represent the real-world behaviour of on-site generation and storage systems.

HEM-TP-18: PV Generation and Self-Consumption

HEM-TP-18 is the technical paper that defines how the Home Energy Model handles on-site photovoltaic generation and electricity self-consumption. Published as part of the HEM technical documentation on GOV.UK, this module sits within the electricity balance step of the core calculation loop. At each half-hourly timestep, after space heating, cooling, and hot water demands have been determined, HEM-TP-18 calculates how much electricity the PV array generates and how that generation interacts with household demand and any battery storage present.

The module draws on solar irradiance data already computed by the external conditions module (HEM-TP-03) and the solar gains module (HEM-TP-08), which calculates direct and diffuse irradiance on surfaces of any orientation and tilt using BS EN ISO 52010-1:2017. HEM-TP-18 then applies panel-specific and inverter-specific parameters to convert irradiance into usable AC electricity.

Half-Hourly PV Generation Modelling

At the heart of HEM's PV modelling is a half-hourly generation calculation that accounts for the actual solar conditions at each timestep. This contrasts sharply with SAP, which estimates an annual PV yield and distributes it across months using fixed proportions.

Irradiance Calculation

For each 30-minute period, HEM determines the total solar irradiance incident on the plane of the PV array. This calculation considers:

• Direct (beam) irradiance: sunlight arriving directly from the sun, dependent on the solar altitude, azimuth, and the angle of incidence on the panel surface
• Diffuse irradiance: sunlight scattered by the atmosphere and clouds, arriving from the whole sky hemisphere
• Ground-reflected irradiance: sunlight reflected from surrounding surfaces onto the panel, calculated using a ground reflectance (albedo) factor

The irradiance data comes from the hourly weather file (in CIBSE TRY or EPW format). The values are not interpolated: the hourly figure applies to both half-hourly timesteps within that hour. The sun position (altitude and azimuth) is calculated geometrically for the site latitude, longitude, and the date and time of each timestep.

Panel Orientation and Tilt Angle

The orientation (azimuth) and tilt angle of the PV array directly affect how much irradiance reaches the panel surface. HEM models this by computing the angle of incidence between the sun's position and the panel normal at each timestep:

• Orientation (azimuth): measured as degrees from due south (0° = south, 90° = west, −90° = east). South-facing arrays in the UK receive the greatest total annual irradiance, but east–west split arrays can improve self-consumption by spreading generation across morning and afternoon periods.
• Tilt angle: measured from horizontal. In the UK, a tilt of approximately 30–40° maximises annual yield for south-facing panels. Lower tilts favour summer generation; steeper tilts favour winter performance.
• Multiple arrays: HEM can model separate PV arrays with different orientations and tilts on the same dwelling (e.g. panels on both east-facing and west-facing roof slopes), calculating generation independently for each array and summing the results.

Shading Effects

Shading can significantly reduce PV output. HEM accounts for shading through overshading factors that reduce the irradiance reaching the panel surface at each timestep. The shading model considers:

• Horizon shading: obstructions on the distant horizon (hills, treelines) that block direct beam irradiance at low solar altitudes, particularly relevant during winter months
• Nearby obstructions: adjacent buildings, chimneys, dormers, or trees that cast shadows on the array. Because these shadows move throughout the day as the sun's position changes, the half-hourly resolution captures the time-varying nature of shading far more accurately than a fixed annual factor.
• Array geometry: the shading calculation uses the array's height, width and base height (the height of its lowest part above ground), so where the panels sit on the roof affects how much of the obstruction matters

The inputs are geometric rather than categorical. Distant objects are described by splitting the ground plane into segments and giving each one an object height and distance; nearby obstructions are described as overhangs and fins. There is no equivalent of SAP's overshading categories.

Inverter Specifications and Efficiency

The inverter converts DC electricity from the PV panels into AC electricity for household use or grid export. Inverter efficiency is not constant; it varies with the load on the inverter relative to its rated capacity. HEM models this relationship explicitly.

Inverter Efficiency Modelling

At each timestep, HEM applies the inverter efficiency to the DC output from the panels to determine the usable AC electricity. The key parameters are:

• Maximum power, both sides: the maximum input power on the DC (panel) side and the maximum output power on the AC (mains) side. Both are required inputs and both are applied as limits, so an array that overruns the inverter is clipped.
• Generic efficiency curves: efficiency is calculated at each timestep from the panel output relative to the inverter's maximum input power. HEM currently carries two curves, one for each inverter type, rather than product-specific data.
• Part-load behaviour: those curves capture the sharp fall in efficiency at low loads (early morning, late afternoon, heavy cloud), which a single headline efficiency figure would hide.
• Peak efficiency already removed: the system performance factors carry an implicit inverter efficiency of 0.972, which HEM divides out before applying the timestep calculation, so inverter losses are not counted twice.

String Inverters vs Micro-Inverters

HEM recognises two inverter types: string and optimised (also described as micro). With a string inverter, a single unit handles the array or a string of panels, and shading on any one panel can drag down the whole string. With an optimised arrangement, that coupling is much weaker. The type drives a separate efficiency factor for partial overshading, applied on top of the reduction in incident radiation, so the distinction only matters where the array is actually shaded.

Self-Consumption Calculation

Self-consumption (the proportion of PV-generated electricity that is used directly within the dwelling rather than exported to the grid) is one of the most important metrics that HEM can calculate at this resolution. At each half-hourly timestep, HEM takes the ratio of PV supply to electricity demand and reads a self-consumption factor from an empirical curve fitted to monitored UK dwellings, capped at 1. Generated electricity is then allocated in priority order:

• Instantaneous demand in the dwelling takes the first share, set by the self-consumption factor
• Battery storage takes the surplus next, if a battery is present and not full
• A PV diverter takes what is left after that, then anything remaining is exported. An input option can put the diverter ahead of the battery

The curve matters. Even when generation exactly matches demand across a timestep, the modelled self-consumption factor is less than 1, because supply and demand do not line up moment to moment within the half hour. A straight comparison of the two totals would overstate how much of the generation is actually used.

The household electricity demand at each timestep includes all electrical loads modelled by HEM: the heat pump compressor and controls, MVHR fan energy, lighting, cooking, appliances, and any electric immersion heater top-up for hot water. Because heat pump electricity consumption is typically the largest single electrical load in a Future Homes Standard dwelling, the timing interaction between PV generation and heat pump demand is critical.

Interaction with Heat Pump Electricity Consumption

The interaction between PV generation and heat pump demand is one of the most consequential modelling improvements in HEM. In winter, when PV output is low but heating demand is high, the heat pump draws heavily from the grid. In spring and autumn, moderate PV generation may partially offset the heat pump's consumption during daytime heating periods. In summer, space heating demand drops to near zero, but hot water heating via the heat pump (or an immersion diverter) can absorb some surplus PV generation.

HEM resolves all of these interactions at half-hourly intervals. The heat pump module (HEM-TP-12) calculates the electricity consumed by the heat pump at each timestep based on the heating load and the coefficient of performance (COP) at that moment. The PV module then determines whether that consumption can be met by on-site generation, battery discharge, or grid import.

Battery Storage Modelling

SAP 10.2 already recognises a battery, but only as a single coefficient: usable capacity, capped at 15 kWh, shifts the monthly self-use factor. HEM replaces that with a simulated battery, charged and discharged timestep by timestep against the actual pattern of generation and demand.

Charge and Discharge Behaviour

At each half-hourly timestep, HEM runs a charge/discharge algorithm for the battery:

1. Surplus PV available? If PV generation exceeds household demand and the battery is not fully charged, the surplus is directed to the battery at a rate up to the maximum charge power
2. Demand exceeds PV? If household demand exceeds PV generation and the battery holds charge, it discharges to meet the shortfall at a rate up to the maximum discharge power
3. Efficiency losses applied: charging and discharging each carry the square root of the round-trip efficiency, so the loss is split evenly between the two directions
4. State of charge updated: the charge level is carried forward to the next timestep, held between zero and the maximum capacity, giving a continuous simulation across the year

Charging from PV may overlap with discharging within the same timestep, which HEM resolves as a net energy flow. Charging from the grid is different: where the input file allows it and a threshold price is set, the battery can charge from the grid, but it cannot discharge at the same time, since the energy could be used from the grid directly.

Battery Specification Parameters

HEM requires the following data for each battery system:

• Maximum capacity (kWh): the ceiling on the charge the battery can hold
• Maximum charge rate (kW): the fastest rate at which the battery can accept energy
• Minimum charge rate (kW): the rate below which charging does not take place
• Maximum discharge rate (kW): the fastest rate at which the battery can release energy
• Round-trip efficiency (%): the proportion of energy recovered after a full charge/discharge cycle, accounting for conversion losses

Two adjustments are then applied to the stated capacity. Age reduces it by 4% a year, based on manufacturer guarantees of 60% remaining capacity after ten years. A battery sited outside the heated space, in a garage for instance, has its capacity reduced further by a factor that depends on the external temperature, reaching no reduction at 20°C and above.

Grid Import/Export Balance

After resolving PV generation, self-consumption, and battery charge/discharge at each timestep, HEM calculates the net grid interaction:

• Grid import: electricity drawn from the national grid when household demand exceeds the combined output of PV generation and battery discharge
• Grid export: surplus electricity sent to the grid when PV generation exceeds household demand and the battery is fully charged (or no battery is present)

The annual totals for grid import and export are key outputs of the HEM calculation. For FHS compliance, the Target Emission Rate (TER) and Target Primary Energy Rate (TPER) apply forward-looking emission and primary energy factors for grid electricity (0.086 kgCO2e/kWh and 1.969 kWhPE/kWh respectively, based on the projected 2025–2029 grid mix). A single annual factor is used rather than time-varying rates. Exported electricity receives a credit that partially offsets the carbon and primary energy of imported electricity.

This granular import/export accounting is beyond SAP's monthly resolution: SAP splits each month's PV output between use and export through a single factor, without knowing when within the month either happened.

Time-of-Use Tariff Implications

HEM's half-hourly resolution opens up a significant new capability: meaningful time-of-use tariff analysis. Because the model knows exactly when electricity is imported and exported, it can distinguish between:

• Cheap off-peak imports: electricity drawn overnight or during low-demand periods when unit rates are lowest (e.g. under Octopus Go, Intelligent Octopus, or Economy 7 tariffs)
• Expensive peak imports: electricity drawn during late afternoon and evening demand peaks when unit rates are highest
• Export value timing: the value of exported electricity varies depending on when it reaches the grid, with potential for higher payments during peak periods under emerging agile export tariffs

While the FHS compliance wrapper currently applies standardised assumptions rather than real tariff data, the half-hourly output from HEM provides the foundation for future policy that could credit smart battery operation, demand shifting, and tariff optimisation. This is directly relevant to the government's broader smart energy strategy and the rollout of smart meters.

SAP's monthly resolution makes time-of-use analysis meaningless: it cannot distinguish between electricity consumed at 2am and electricity consumed at 6pm.

SAP vs HEM: Solar PV Modelling Comparison

The table below summarises the fundamental differences in how SAP and HEM handle solar PV and battery storage:

For the full technical comparison across all HEM modules, see the HEM Technical Reference. For a non-technical overview of the differences, see SAP vs HEM.

FHS Mandatory Solar Requirement

Solar PV is not merely a compliance option under the Future Homes Standard; it is a functional requirement (Requirement L3). In the March 2026 consultation response, the government confirmed that on-site renewable electricity generation is mandatory for new-build homes under the FHS, moving beyond the original consultation position where solar was included in the notional building specification rather than being formally required.

40% Floor Area Coverage Target

The solar requirement targets PV coverage equivalent to approximately 40% of the ground floor area of the dwelling, using panels with an efficiency of at least 0.22 kWp/m² (per AD L1 Equation 5.1). For a typical three-bedroom semi-detached house with a ground floor area of around 45 m², this equates to roughly 18 m² of PV panels, approximately 4.0 kWp.

The requirement is written as an output test, not an area rule, and AD L1 gives two ways to satisfy it:

• The notional array: an annual output at least equal to that of an array sized at 40% of the ground floor area, oriented south-east to south-west, pitched at 45 degrees and unshaded
• The practicable roof: an annual output at least equal to that of an array covering the reasonably practicable roof area at 0.22 kWp/m²

For a building containing dwellings, the ground floor area is the total floor area of all spaces divided by the number of storeys, and an individual dwelling's share is apportioned from that by floor area. The generated electricity must be available to residents, which in a block can mean individual dwellings and communal spaces.

Exemptions and Exceptional Circumstances

AD L1 paragraph 5.77 recognises one exceptional circumstance: insufficient roof area. For a dwellinghouse, that means demonstrating that the roof cannot take an array capable of generating 720 kWh a year. For a building containing dwellings, the same test applies per dwelling, divided by the number of storeys. Appendix B of AD L1 sets out how the reasonably practicable roof area is established for different roof types.

Falling short is not a free pass. Where the standard cannot be met, paragraph 5.76 requires the design to be modified to maximise output: higher performance panels, alternative panel orientations, and moving roof vents or architectural features to make room for more panels. Any array that does not meet the standard is highlighted in the BREL report, and evidence for the roof area available has to go to the building control body.

Product Data Requirements

HEM's accuracy depends on the quality of the product data provided for PV panels, inverters, and battery systems. The model requires significantly more detailed specifications than SAP:

Panel Specifications

• Rated power output (Wp): the peak watt rating under Standard Test Conditions (STC: 1000 W/m² irradiance, 25°C cell temperature)
• Panel area (m²): used to determine the total array area and coverage relative to the FHS target
• Ventilation of the modules: unventilated, moderately ventilated, strongly or forced ventilated, or free-standing. This sets the system performance factor, which runs from 0.81 to 0.87 in HEM, uplifted from the values in the BS EN standard using monitored UK system data
• Array geometry: height, width and base height, needed for the shading calculation, alongside orientation and tilt

Two parameters a specifier might expect are not asked for. Module temperature coefficient and Normal Operating Cell Temperature are accounted for only implicitly, inside the annual system performance factor. HEM-TP-18 lists both, along with degradation rate and product-specific module efficiency, as candidates for explicit treatment in future.

Inverter Specifications

• Maximum power on the DC (panel) side
• Maximum power on the AC (mains) side
• Inverter type: string or optimised (micro). There is no third category, and no efficiency input. Both maxima are applied as limits, and the efficiency curve follows from the type

Battery Specifications

• Maximum capacity (kWh)
• Maximum and minimum charge rate (kW)
• Maximum discharge rate (kW)
• Round-trip efficiency (%)
• Whether the battery sits inside or outside the heated space

The Product Characteristics Database (PCDB), currently being revised for HEM, will be the primary source for verified product data. For PV today, though, much of the detail is generic rather than product-specific: the system performance factor comes from a table indexed by module ventilation, and inverter efficiency from one of two generic curves. Replacing those generic curves with product-specific data through the database is listed as a future development, which would give the more efficient models something to show for it.

Emerging Solar Policy

At a national level, the government's Solar Roadmap sets out how to reach 45 to 47 GW of installed solar capacity by 2030. This ambition reinforces the growing importance of accurate solar modelling in tools like HEM. As domestic PV installations continue to grow, the ability to model generation, self-consumption, and battery interactions at half-hourly resolution becomes essential for both regulatory compliance and realistic energy cost projections.

Practical Implications for Assessors and Designers

The shift from SAP's simple annual yield calculation to HEM's half-hourly PV and battery modelling has significant practical consequences:

System Design Optimisation

Because HEM rewards genuine self-consumption and battery performance, system designers can now optimise PV and battery configurations with confidence that the compliance model will reflect real-world benefits. Key design decisions that HEM properly values include:

• East–west split arrays: spreading generation across morning and evening periods to improve self-consumption, even though total annual yield is slightly lower than a south-facing array
• Right-sized battery capacity: matching battery size to the gap between daytime generation and evening demand, rather than simply installing the largest available system
• Inverter sizing: selecting an inverter rating that balances peak generation capture against efficiency at typical operating loads
• Heat pump scheduling: where controls permit, running the heat pump preferentially during high-generation periods to absorb PV output directly rather than exporting it

Data Collection Requirements

SAP assessors transitioning to HEM will need to collect considerably more data for PV and battery systems. Under SAP, a PV system requires little more than the array size, orientation, and tilt. Under HEM, assessors must also specify the array geometry needed for shading (height, width and base height), the module ventilation category, the inverter's DC and AC power limits and type, and the full battery parameter set where a battery is present.

Some respondents to the Future Homes Standard consultation reported HEM assessments taking around three times as long as the equivalent SAP assessment. The additional PV and battery data contributes to this increase. Setting up a project data room from the design stage, with all product specifications readily available, is strongly recommended.

Frequently Asked Questions

Related Pages