Rolling Out Solar Across a Property Portfolio

Most landlords approach solar one building at a time — a feasibility study here, a quote there, a decision deferred until a lease event forces it. For a single owner-occupied unit that is reasonable. For a portfolio of ten, fifty or two hundred assets it is the most expensive way to do the work: every building re-tenders the same survey, re-negotiates the same finance, and re-invents the same legal structure, while the estate’s aggregate EPC, GRESB score and CRREM stranding position drift with no one accountable for them as a set. A portfolio programme replaces that with one standardised pipeline — survey to switch-on — run across every asset on a common framework. This guide sets out how that programme is structured, funded and reported, and where the legal and tax detail bites when solar sits on assets that are bought, sold and refinanced.

This is the pillar for portfolio owners. If you hold a single building, the route-by-route mechanics in solving the split incentive and the asset-led case in the green premium will serve you better. If you manage a fund or a multi-asset estate, read on.

Why a programme beats ad-hoc building-by-building

A programme wins because it standardises the three things that are slow and expensive to repeat: feasibility, design and finance. Done once as a framework and applied across the estate, each becomes a unit cost that falls with volume rather than a fresh project every time.

Standardised feasibility and design framework

The first move is a single estate-wide screen, not a building-by-building survey. Each asset is scored against a common set of inputs — roof area and condition, structural loading headroom (assessed to [BS EN 1991](https://www.gov.uk/ MEES guidance) loading principles), current EPC band, metered or estimated daytime load profile, DNO grid headroom at the connection, lease structure and remaining roof membrane life. That produces a ranked pipeline rather than a pile of one-off reports. A standard design template — module layout logic, inverter sizing rules, a default battery decision tree, a fixed G99 application pack — is then applied to each qualifying roof, so the engineering effort per building drops sharply after the first cohort.

Prioritisation by MEES risk, roof condition and DNO headroom

Not every building should go first, and the sequencing is where a programme earns its keep. The standard prioritisation logic weighs three things:

• MEES risk. It has been unlawful to let a commercial property in England and Wales below EPC E since 1 April 2023, for sitting tenants as well as new lets. Assets sitting at F or G, or at a marginal E with a lease event approaching, move to the front — solar typically lifts a commercial EPC by one to three bands (never assume a specific jump), and on a borderline asset that can be the difference between lettable and stranded. The proposed EPC B by 2031 standard for privately-rented non-domestic buildings over 1,000 m² (Government interim response, 18 June 2026) is a proposal subject to secondary legislation, not yet law — but it shapes which assets are worth doing early.
• Roof condition. A roof with under fifteen years of membrane life left should be re-roofed and solar-fitted in one project, not solar-fitted now and disturbed later. Roof lifecycle is a sequencing input, not an afterthought.
• DNO headroom. Above roughly 50kW the G99 grid connection is now the real delivery bottleneck, not planning. Sites with existing headroom can be switched on quickly; sites needing reinforcement go into a longer track so a single constrained connection does not stall the whole programme.

The full background to the MEES position is in the MEES and EPC guide, and the grid-connection mechanics in the planning and grid guide.

Framework funding across the estate

The funding question for a portfolio is not “how do we pay for this building” but “what mix of capital pays for the estate” — and the answer is usually a blend, not a single source. A framework lets each asset draw on the route that fits its tenure, covenant and balance-sheet position while the programme is procured and managed as one.

A few points matter when blending these at portfolio scale. Solar PV is a special rate (integral features) asset, so the headline 100% first-year relief comes from the Annual Investment Allowance — £1m and permanent — not from full expensing, which only delivers the 50% First-Year Allowance on solar and is unavailable on any asset bought to lease. That single rule pushes most landlord and roof-lease structures onto AIA plus the 6% writing-down allowance, and it is why the funding mix has to be set asset-by-asset rather than estate-wide by default. The detail sits in capital allowances and funding for owners.

Third-party-funded roof leases deserve a specific note in a portfolio context. They take the kit off your balance sheet and pay you rent, which is attractive across let assets where you do not want to deploy capital — but a roof lease is a registrable interest in land. It carries SDLT and Land Registry consequences, needs mortgagee and insurer consent on every affected title, and typically runs twenty to thirty years. Negotiated once as a framework with one counterparty across many roofs, that legal cost is amortised; negotiated building-by-building it is punitive. The trade-offs between this, a PPA and a licence are set out in roof lease vs PPA vs licence.

One reporting line: EPC, GRESB and CRREM

The strategic prize of a portfolio programme is that it moves three estate-level metrics at once, on a single evidence base, reported through one line rather than reconstructed from scattered building files.

Estate-wide EPC and stranding position

Around 83% of commercial buildings across seven major UK cities sit below EPC B (BPF, October 2025), Savills puts roughly 185 million sq ft of UK retail at risk of being unlettable, and CBRE finds about 58% of Central London offices below B. Against that backdrop, an estate’s aggregate EPC profile is a risk register, not a compliance checkbox. A rollout that lifts the worst assets first improves the distribution measurably — and because the programme captures each asset’s before-and-after band on a common template, you get a defensible estate-wide EPC trajectory rather than a guess.

GRESB performance and the CRREM pathway

On-site solar is Scope 2 abatement, and at portfolio level it feeds directly into the disclosure frameworks investors now price:

• GRESB. The 2025 assessment covered 1,002 fund managers, 2,382 assessments and 150+ institutional investors. A standardised rollout generates the consistent, evidenced data points GRESB scores reward — generation, consumption, emissions reduction — across every asset on the same basis.
• CRREM. The 1.5°C asset-level decarbonisation pathways show where each building strands and when. Sequencing solar by stranding date keeps assets above their CRREM line for longer, and the programme documents the intervention against the pathway.
• SFDR, TCFD and SBTi Buildings. Funds disclosing under SFDR Article 8/9, reporting to TCFD or setting SBTi Buildings targets need auditable, asset-level energy and emissions data. A building-by-building approach produces inconsistent evidence; a programme produces one dataset.

The full reporting architecture — how solar evidence maps onto GRESB, CRREM, SFDR and TCFD — is covered in ESG, GRESB and net zero for funds.

s.198 fixtures elections on every transaction

Once solar is installed, it becomes a fixture — and that has a permanent consequence every time an asset is bought or sold inside the portfolio. A capital allowances s.198 election is needed when buying or selling let property with existing solar, to fix the value attributed to the plant and machinery (including the solar PV as an integral feature) for both parties. Get it wrong and allowances can be lost entirely.

For a portfolio that trades assets, this turns from a one-off into a standing process. Every acquisition of a building with solar should screen for an existing election and the allowances pool; every disposal should set the s.198 figure deliberately rather than by default. Built into the programme’s transaction checklist, it protects the tax position across the whole estate. Treated as an afterthought at each deal, it leaks value. The mechanics, including the interaction with HMRC’s capital allowances treatment, are in capital allowances and funding for owners.

Covenant, balance sheet and phasing

Two practical constraints shape the order of a rollout as much as MEES or grid: tenant covenant strength and the owner’s balance-sheet capacity — and both argue for phasing rather than a single estate-wide spend.

Covenant strength matters because it determines which funding route an asset can carry. A let building with a strong tenant covenant supports a landlord-to-tenant PPA or a roof lease cleanly; a weaker covenant may make third-party funders cautious and push the asset toward common-parts capex (lifts, lighting, HVAC, car park and EV — where the landlord both pays and consumes, with no split incentive to solve). The split-incentive routes are set out in full in solving the split incentive.

Balance-sheet capacity sets the pace. Few owners want to deploy capital across an entire estate in one year, and they do not need to — a phased programme deploys capex where the returns are strongest and the MEES risk highest, uses asset finance to spread the next cohort, and hands the let, weaker-covenant assets to a third-party roof lease that needs no owner capital at all. The phasing also respects the grid: G99 connections above 50kW are sequenced so reinforcement works on one site do not block switch-on across the rest. The result is a rollout that lands on its own cashflow, lifts the estate’s worst MEES and stranding risks first, and keeps a single connection bottleneck from holding up the programme.

A typical phased shape:

1. Phase 1 — urgent and easy. Sub-E and marginal-E assets with existing grid headroom and good roofs. Capex or asset finance. Immediate compliance and EPC benefit.
2. Phase 2 — strong returns. High-load owner-occupied and single-tenant assets where self-consumption (the single biggest return driver — 50–70% on a daytime-shifted site, 60–80% with battery) makes the case strongest.
3. Phase 3 — let and constrained. Multi-let and weaker-covenant assets via PPA or roof lease, and sites needing DNO reinforcement, run on a longer track.

Frequently asked questions

How many buildings do you need before a programme makes sense?

There is no hard threshold, but the economics of standardisation usually start paying back at around five to ten assets. Below that, the framework set-up cost — common feasibility template, master funding terms, a single roof-lease counterparty — is hard to amortise, and a building-by-building approach is fine. Above it, the per-building cost of survey, design and finance falls with each cohort, and the estate-level EPC, GRESB and CRREM reporting becomes worth running as one line. The deciding factor is less the count than whether you are managing the estate’s aggregate stranding risk as a portfolio or as a stack of separate buildings.

Does a rollout have to be funded one way across the whole estate?

No — and it usually should not be. The point of a framework is that each asset draws on the funding route that fits its tenure and covenant: capex on owner-occupied high-load buildings, asset finance to spread the next cohort, and a third-party roof lease on let assets where you want rent rather than capital deployment. The tax treatment differs by route — AIA-backed 100% relief on capex, 6% writing-down allowances on financed kit, rental income with no allowances on a roof lease — so the mix is set asset-by-asset and procured as one programme.

Will solar guarantee our buildings hit the proposed EPC B standard?

No — and any provider promising a specific band jump is overstating it. Solar typically lifts a commercial EPC by one to three bands, but the actual movement depends on the building’s baseline fabric, services and assessment inputs. The proposed EPC B by 2031 standard (for privately-rented non-domestic buildings over 1,000 m², from the Government’s 18 June 2026 interim response) is a proposal subject to secondary legislation, not current law. The only binding legal floor today is EPC E. A rollout should be sized to move each asset’s measured band as far as the modelling supports, sequenced by which assets carry the most MEES and stranding risk — not sold on a headline band promise.

Who manages the programme, and what does the owner actually have to do?

The owner sets the strategy and approves the pipeline; the programme handles survey, design, funding, grid and delivery on a single line of accountability. In practice that means one feasibility screen across the estate, one ranked and sequenced pipeline for sign-off, one funding framework, and consolidated reporting that feeds straight into your GRESB submission, CRREM tracking and SFDR or TCFD disclosure. The owner’s job is the decisions that need ownership-level judgement — capital allocation, lease structure, which assets to prioritise — not re-running the same procurement at every building. To scope a programme across your estate, request a portfolio assessment or read the property-type detail for commercial property portfolios.