Project
overview

Context

The UK’s social housing sector faces a pressing challenge: how to deliver energy efficiency improvements at scale and at a cost that is financially and socially sustainable. As part of a consultation on minimum energy efficiency standards (MEES), the UK Government has proposed a deadline for all social housing properties to achieve an Energy Performance Certificate (EPC) rating of C - according to reformed EPC metrics - or above by 2030.¹ There are currently approximately 1.2 million social homes below this threshold.

Yet housing associations face substantial financial challenges, with competing demands on their finances. This includes requirements for delivering new housing supply, meeting health and safety standards, undertaking remediation works, and funding energy efficiency improvements. Furthermore, evidence suggests that many housing associations encounter significant uncertainty around the true costs of retrofit and the ability of existing assessment tools to predict a building’s thermal performance. There is therefore potential for a more efficient use of scarce capital where the true costs and scope of retrofit are known and are targeted where most needed.

Context

Purpose

To support the industry’s exploration of potential solutions to this challenge, Lloyds sponsored a pilot project with Bromford Flagship^a, a housing association, and Senze, a technology provider.

The pilot tested the use of monitoring technology in a sample of 121 homes to measure actual thermal performance and compare this with modelled EPC data. The technology platform then used that real-time property-level data to suggest which properties to target for retrofit, and the specific interventions likely to be most effective in upgrading the performance of each individual property to the standard of an EPC C, on a granular, room-by-room basis. This service is delivered at a typical cost of £1,500 per property.

The project then assessed the recommended retrofit works and examined the costs and potential implications of delivering them against Bromford Flagship’s expected spend based on a traditional EPC-led approach.

The pilot set out to test a central hypothesis: how can sensor-generated measured data provide the intelligent insights needed to make better retrofit decisions? This question is considered relative to a traditional EPC-led approach while balancing environmental, social and economic outcomes in the context of the commercial and regulatory constraints that the social housing sector must work within.

Many of the findings are also likely to be of relevance to other types of property, including owner-occupied housing, other residential tenures for rent, and commercial real estate, particularly in older stock requiring retrofitting to meet new environmental standards.

_{^aBromford merged with Flagship on 28 February 2025 to create Bromford Flagship, an 80,000-home housing association. For this project, we primarily interacted with Bromford, and most of the homes included in the pilot sample were originally from Bromford’s portfolio, with a small proportion from Flagship’s portfolio.}

Purpose

To support the industry’s exploration of potential solutions to this challenge, Lloyds sponsored a pilot project with Bromford Flagship^a, a housing association, and Senze, a technology provider.

This report

The Good Economy (TGE) has been commissioned to author this report, to provide an independent view on the expected environmental, social and economic impacts resulting from the pilot programme, and to consider the potential implications of wider adoption of this type of technology.

As an input to this report, Birmingham City University has been commissioned to verify Senze’s method to collect the property-level data and to recommend targeted retrofit interventions. This is intended to provide assurance regarding the technical robustness of the technology used.

As of October 2025, the technology has been used to collect baseline energy and thermal performance data on a sample portfolio of homes, and to provide recommendations on works and costings to upgrade their performance. The works themselves have not yet been carried out. Therefore, this report is forward-looking.

It provides an overview of the benefits we expect to arise from this programme, assuming the recommendations from the technology deployment are implemented, rather than an assessment of the actual effect of implementing its recommendations.

This report provides TGE’s headline findings on the pilot study. Further information is provided in The Science Behind the Insight document which covers contextual background, detailed findings, key assumptions, challenges and limitations, and technical commentary on the expected outcomes and implications and the methodologies used.

This report

Key project partners

Lloyds is the largest lender to the UK social housing sector and has provided over £20 billion in finance to the sector since 2018 through commercial lending and deal facilitation, enabling more homes to be built.

Lloyds has also committed to lend up to £500m to support social housing retrofit, partially guaranteed by the National Wealth Fund (NWF) as part of a £1.3bn social housing retrofit guarantee scheme established by the NWF. As a major lender to social housing as well as the real estate sector more generally and as a signatory to the Net Zero Banker’s Alliance, Lloyds is keen to explore solutions that will support the industry to meet its transition goals.

Bromford Flagship is a housing association operating across East, Central and South West England that owns and manages over 80,000 homes. Bromford Flagship’s involvement in the Senze and Lloyds data-led retrofit pilot project was driven by a commitment to achieving both immediate energy efficiency improvements and long-term progress towards Net Zero 2050. This pilot provides an opportunity to test innovative, data-driven approaches that move beyond traditional EPC modelling, offering more accurate insight into actual building performance and energy use. By harnessing real-time data, Bromford Flagship’s aims are to ensure that retrofit measures deliver measurable reductions in carbon emissions and resident energy costs. The anticipated outcomes include the development of an evidence-based framework to inform future retrofit programmes, improved tenant satisfaction through enhanced comfort and affordability, and the creation of scalable solutions that can be applied across the wider housing sector.

Senze is the technology provider whose sensors and technology have been used to measure the performance of the pilot properties and provide recommendations for retrofit works. Ultimately, this pilot supports Senze’s mission to explore technological solutions to facilitate the decarbonisation of the built environment.

Figure 1: Stakeholder Diagram

Lloyds

_The_sponsor_{of the retrofit pilot programme and a major lender to the UK social housing sector}

_{Produce independent report}

Bromford Flagship

_The_{Housing Association}_{that owns the portfolio of 121 social housing properties that were subject of the retrofit pilot}

↓ Information ↓

The Good Economy

_The_{author of a third-party report}_{, commissioned by Lloyds, exploring and providing an independent view on the expected economic, social and environmental impacts resulting from the pilot programme}

↑ Information ↑

Birmingham City University

_The_{third-party verifier}_{, responsible for reviewing and assessing the robustness of Senze's method used to collect the property-level data, and verifying the data collected}

CFP Green Buildings

_{Provider of independent unit cost data for retrofit interventions via the CFP Green Buildings Tool}

_{Collecting performance data}

_Information

_{Verifying data and method}

_{Pricing baselines}

Senze

_The_{retrofitting platform}_{whose sensors and data dashboard were used to measure the performance of the properties and provide recommendations for retrofit works}

University of Salford report

_{University of Salford’s Energy House 2.0 research facility was used to validate the efficacy of Senze’s Live Thermal Measurement Algorithm}

_{Commissioned by Senze, separately to this project}

Summary findings

Though based on a small and deliberately skewed sample, the findings from the pilot study are significant:

The data gathered validates the widely shared view that EPCs are often inaccurate in their estimations of thermal performance and provides insight as to exactly how these inaccuracies manifest. Significant discrepancies were observed between individual homes' modelled EPCs and their measured thermal performance (see chapter 4 for an overview of the difference between a 'modelled' vs a 'measured' approach). The average thermal performance gap was 25%, with 13 properties (11% of the pilot sample) experiencing a level of heat loss which was more than double their modelled predictions. This raises important questions about the reliability of EPCs as a decision-making tool for retrofit investment, a point which has been highlighted by other organisations, notably the Climate Change Committee.²

Analysis of thermal performance by property age revealed that older properties in the pilot sample generally outperformed, or aligned more closely with, modelled predictions compared with newer homes. Although the sample size, particularly for newer properties, was limited, the finding warrants further investigation. If confirmed at scale, it may indicate that older properties require fewer interventions than expected, while newer properties may require greater attention.

The monitoring technology appears to enable more accurate scoping of retrofit works, avoiding unnecessary interventions, vs what had been recommended based on a modelled approach using EPC data. This approach enables more efficient use of resources through targeted delivery tailored to each property's needs, rather than applying a blanket fabric-first model. Traditional modelled approaches can result in unnecessary or even harmful insulation measures and the installation of oversized heating systems. For example, the measured data showed that, for the majority of properties, a 10kW heat pump would be oversized and so not required based on peak heat load.

There are potentially significant economic outcomes, including cost savings, to be realised through adoption of this type of technology-enabled, measurement-based approach. 59% of properties in the pilot sample were assessed to require retrofit based on both measured and modelled data. At these properties, the measured data showed that, on average, works could be delivered in a more targeted manner, and at a lower cost to achieve an EPC C, than had been planned by Bromford Flagship based on a modelled approach. The data suggests that this could lead to an average saving of £12,500 per property for the properties in this group, based purely on the cost of the interventions themselves. This could rise to over £27,500 per property if also factoring in potential savings through a more commercial procurement approach, based on independent costing data for retrofit interventions provided by CFP's Green Buildings Tool.^b

For a further 7% of properties in the pilot sample, measured data suggested these homes did not require any interventions to achieve a level of thermal performance equivalent to an EPC C or better. These properties had been earmarked for retrofit, with Bromford Flagship having planned the works and set aside funding. Given the measured data suggests that these properties do not require any interventions, deploying this type of technology could result in an average saving of over £41,000 per home for the properties in this group if it was deemed that these interventions were not required.^c

There is also the potential for an increase in costs at properties where homes are shown to be significantly underperforming their modelled predictions following retrofit based on a traditional approach. 13% of the properties in the portfolio were shown to require interventions to bring them up to the standard of an EPC C equivalent, even though their stated EPC rating was already a C or better. Where data was available, Bromford Flagship had already spent an average of £27,000 on retrofit at these properties, yet the measured data demonstrates a risk that these retrofit resources have not delivered the expected improvement in performance as yet. Using the technology to identify this category of properties highlights the risk of ineffective spending based on current methods.

Deployment of this type of technology offers the potential for social impact through targeting properties and retrofit interventions that will be most effective in addressing home comfort issues in addition to improving energy performance. This could include, for example, targeting properties where measured data shows a material risk of damp and mould, and/or ensuring that homes are suitably insulated. In addition, interventions can be targeted at properties that would otherwise have fallen out of scope of a typical modelled retrofit programme, but where measured data shows they underperform against their modelled predictions.

_{^bSee more details on the cost comparison methodology underpinning this calculation in the ‘Economic benefits’ section of this report in}_{chapter 7}_{, and in 'Appendix 7: Potential Outcomes' in The Science Behind the Insight.'^cNote this cost-saving would be contingent on the EPC regime being updated to accept measured inputs – this point is currently being considered by Government within its}_consultation_{to reform the energy performance of buildings regime, with a decision expected in 2026.}

A key consideration for any future deployment of this technology is the inherent trade-offs between environmental, social and economic outcomes. In this pilot, measured EPC C compliance was used as the guiding principle for identifying the most effective works within the current policy framework. However, due to the nature of EPCs as a cost-based measure, this can result in recommended interventions that do not maximise carbon reduction. The technology used in the pilot can flexibly accommodate these different priorities, but this highlights the trade-offs that exist and which housing associations must manage in their approach to retrofit given their overriding social purpose.

The nature of the technology tested in this pilot allows for a low-risk, agile approach to implementation. The relatively low cost of deploying technology in each home compared with the average scale of potential savings allows for an incremental rollout, prioritising homes with the most urgent need, rather than committing significant capital up front.

Whilst the size of the pilot is small, it signals that the implications of the use of the technology for the sector could be profound. If results were replicated across different geographies, property types, and portfolios, and then such technology adopted at scale, this approach could unlock significant cost savings and accelerate the decarbonisation of the built environment.

For housing associations, it offers a potential pathway to deliver retrofit more affordably and effectively. For lenders and investors, it could strengthen the investment case and reduce risk exposure. For residents, it provides a potential route to warmer, healthier homes with lower energy bills.

To realise its full potential, the sector will need to invest in further monitoring technology pilots, deepen collaboration across housing, finance and technology partners, and build the evidence base required to drive systemic change. We hope that the findings of this pilot study will help to stimulate conversations within the social housing sector and beyond around the possible benefits and the role that data and technology could play in supporting retrofit design and implementation.