Win Cleaning Contracts When Price Is Only 40 Per Cent

Tenant energy data is the activity data behind Scope 3 Category 13 and a landlord does not hold it by default. This article answers one question. Without sub metering where can a UK landlord get measured tenant energy data and what does each route produce? The answer draws on UK meter statistics and the reports of six UK listed landlords and eight peer reviewed studies. The finding is that two thirds of smaller non domestic meters are already smart or advanced so the interval data largely exists. Coverage is decided by the route used to collect it and not by the lease clause that permits it.

Summary of findings

Most of the meters are in place. The latest government count is 2.1 million smart and advanced meters in smaller non domestic sites in Great Britain. That is 67% of the non domestic meters covered by the rollout [1]. The count covers electricity and gas meters together. Elexon is due to complete the implementation of half hourly settlement for electricity in May 2027 [2] [3].

Access decides coverage. Two landlords with data sharing provisions in their leases report coverage of 99% and 18% on their own definitions. The higher figure is whole building energy data coverage at British Land which pairs its clauses with smart metering on 86% of its portfolio by estimated rental value [4]. The lower figure is occupier data coverage at Hammerson where most leases carry green lease standards and coverage fell from 27% to 18% [5].

A whole building meter still gives a measured total when tenants share nothing. The split between tenants is then an allocation and the rule changes the answer. In a worked example the GHG Protocol occupancy adjustment moves one tenant allocation by 25% [6].

Estimates are the weakest route. A review of 62 non domestic buildings found measured energy use deviated from prediction by +34% on average with a standard deviation of 55% [7]. Analysing one meter to infer individual loads is not a substitute. A field study in three commercial buildings found the method largely failed to identify specific loads [8].

Why tenant energy data matters for Scope 3

Category 13 is often the largest part of a landlord footprint and it is built from data the landlord does not hold. At LondonMetric it was 105114 of the 111434 tCO2e in the greenhouse gas inventory for calendar year 2025. That is 94% of the total. The Category 13 figure was itself extrapolated from 72% actual data received [9]. Sub metering every demise is the obvious fix. It is also slow and costly and depends on access to tenant space.

The Financial Conduct Authority (FCA) final rules in PS26/19 apply the UK Sustainability Reporting Standards to listed companies in scope on a comply or explain basis for accounting periods beginning on or after 1 January 2027. A one year relief covers Scope 3 so Scope 3 disclosures must be addressed for periods beginning on or after 1 January 2028 [10]. UK Sustainability Reporting Standard S2 (UK SRS S2) directs a company measuring Scope 3 to prioritise primary data all else being equal (paragraph B47). It names meter readings and utility bills as sources of primary data (paragraph B48) [11]. A landlord that cannot sub meter needs to know which other routes give measured data and how each will read to an assurance provider.

Definitions

Half hourly data is electricity consumption recorded in half hour intervals and collected automatically from the meter. MPAN is the Meter Point Administration Number which identifies each electricity supply point in Great Britain. Both terms are defined in the Sustainify AI glossary.

Activity data is the physical quantity a calculation is applied to. For tenant energy it is kilowatt hours (kWh). Sub metering means meters installed downstream of the incoming supply so that consumption can be attributed to a demise.

Data quality grade records how a figure was obtained. This article uses three grades. Measured means read from a meter or an actual invoice for the demise. Allocated means a measured building total split by a rule. Estimated data means floor area multiplied by a benchmark. The glossary counts a whole building figure apportioned across occupiers as estimated data. This article grades it separately as allocated because the building total is measured and only the split is assumed.

Net lettable area (NLA) is the floor area a tenant leases and pays rent on. Common parts are the areas a landlord retains and operates. Organisational boundary defines whose emissions a company reports. Whether a demise falls in Category 13 depends on the type of lease and the boundary chosen [12]. A building management system (BMS) runs heating and cooling and other plant and holds operational data for the plant it controls.

Method

Three kinds of source were used. The first was UK statistics on smart and advanced meters from the Department for Energy Security and Net Zero (DESNZ) and the settlement timetable from Elexon and the Market wide Half Hourly Settlement (MHHS) programme. The second was the latest reports of six UK listed landlords that describe how they collect tenant data (Table 2). The third was peer reviewed studies found through Google Scholar on green leases and estimation accuracy and load disaggregation. Each study was checked at its publisher page or its Digital Object Identifier (DOI) record or an open repository copy.

All sources were checked on 6 October 2026. Coverage figures are quoted as each company defines them. The worked example in Finding 3 applies the GHG Protocol allocation formula to an invented building and is labelled as such.

Findings

Finding 1. Two thirds of smaller non domestic meters are already smart or advanced

At the end of June 2026 large suppliers operated 1.3 million smart and advanced meters in smaller non domestic sites. That is 65% of the non domestic meters they operate. Including small suppliers whose latest data is from the end of December 2025 the total is 2.1 million or 67% of all non domestic meters covered by the rollout [1].

The count needs two qualifications. It covers electricity and gas together and only a quarter of non domestic meters are gas meters. The bulletin text does not split the 67% by fuel. It also mixes two meter types. Of the 1.3 million at large suppliers 0.7 million were advanced meters and 0.6 million were smart meters. 89% of those smart meters were operating in smart mode [1]. An advanced meter must at minimum store half hourly electricity and hourly gas data which the supplier can access remotely [13].

Table 1. Smart and advanced meters at smaller non domestic sites in Great Britain

Source. DESNZ statistical bulletin published 27 August 2026 [1] and MHHS programme milestones [3]. Small supplier figures run to December 2025.

Settlement is moving the same way. Elexon is due to complete the implementation of half hourly settlement in May 2027. It says the move means settlement can be based on actual consumption for each half hour in place of estimates or monthly readings [2]. All suppliers must have migrated every MPAN into the new arrangements by 7 May 2027 [3]. Migration does not change the meter. The new arrangements still take in non smart and unmetered supplies [2].

For about two thirds of smaller non domestic meters interval data can therefore already be recorded at the meter. Settlement is a market process and does not give a landlord access. For those meters the barrier is access and not measurement.

Finding 2. Coverage follows the collection route more than the lease clause

The clause is common where a lease is green. Janda and colleagues report an analysis of 26 UK commercial leases. Fifteen contained one or more green clauses and within those data sharing was the most common clause at 80%. Eleven had no discernible green clause even though the sample was skewed toward green buildings. The same paper notes that leases seldom require the parties to share energy data [14]. The Better Buildings Partnership (BBP) Green Lease Toolkit includes a data sharing clause. Under it the tenant authorises the landlord to contact directly any third party that provides supplies to the premises or collects data on them. The tenant must also provide a letter of authority on request [15].

What turns the clause into data is the route behind it. Table 2 sets the route each landlord describes against the coverage it reports.

Table 2. Collection route and reported data coverage

Source. Company reports as cited. Each company defines coverage differently so the figures are not like for like.

Both ends of the range carry clauses. British Land pairs them with smart metering and reports 99% whole building energy data coverage [4]. Hammerson reports green lease standards in the majority of its leases and occupier data coverage of 18%. It plans automated capture in 2026 to improve that coverage [5]. LondonMetric shows the cost of a manual route. Its coverage fell from 80% to 72% which the company attributes mainly to industry changes in how data is collected. The data behind its extrapolation covers 55% of portfolio floor area [9]. SEGRO reports 91% visibility although its standard lease terms give no automatic visibility [16]. This is a comparison of six companies and not a measured effect. It is consistent with the research view that in any tenanted space the opportunity for environmental change is mediated through the lease [19]. The lease permits. It does not collect.

Finding 3. A whole building meter gives a measured total and an allocated split

Where tenants share no data the landlord often still holds the incoming meter. Derwent London takes landlord area as gross internal area minus net lettable area which is broadly the common parts and plant. It calculates landlord area electricity by dividing total building consumption by gross internal area and multiplying by the landlord area. Occupier use is the total minus the landlord share [17]. Where customers have their own utility supplier Workspace estimates their electricity from the energy use of the rest of the building by floor area pro rating [18].

The GHG Protocol gives a formula for buildings that are not sub metered. Energy for a leased space equals its area divided by the product of the building total area and the occupancy rate. The result is multiplied by building total energy use (Formula 8.2) [6]. The Category 13 guidance refers back to that method. Its own example allocates by floor space alone with no occupancy adjustment [12].

Table 3. Worked example of floor area allocation with and without occupancy

Source. Worked example using GHG Protocol Formula 8.2 [6]. The building is invented and is not an observed asset.

The total is measured. The split is a rule. Two defensible rules give answers 25% apart for the same tenant. The rule therefore needs to travel with the figure.

Finding 4. Estimates without meter data can sit far from measured use

The fallback when nothing is metered is floor area multiplied by a benchmark intensity or a model. Research on non domestic buildings shows how far that can drift from what meters record.

Table 4. What peer reviewed studies found about estimates and measurement

Source. Abstracts and open text of the studies cited.

The same review estimated that occupant behaviour can change energy use by 10% to 80% and poor operational practice by 15% to 80% [7]. Occupant behaviour sits with the tenant. A benchmark cannot see it. Estimated data is legitimate but it is a lower data quality grade and it should be calibrated against measured buildings where any exist.

Finding 5. Analysing one meter does not replace a split in multi let buildings

Non intrusive load monitoring infers individual loads from the current and voltage measured at a single supply point. Hart set it out in 1992 as an alternative to intrusive monitoring of each load [22]. Norford and Leeb reported that the technique had worked well in houses but faced more difficult challenges in commercial buildings [23]. Meier and Cautley installed a system in three commercial buildings and found it largely failed to identify specific loads [8].

The practical conclusion is simple. Half hourly data from a tenant MPAN is measured data for that tenant. Half hourly data from a shared incoming meter is measured data for the building and still needs an allocation rule for each tenant.

What this means in practice

Five routes ranked by data quality grade

Table 5. Routes to tenant energy data without sub metering

Source. Classification by this article based on Findings 1 to 5.

Four actions for the next reporting period

1. Build an MPAN register for tenant supplies

List each tenant MPAN against its demise and its lease. A tenant that buys its own electricity has its own supply point even where the landlord has no sub meter. The MPAN stays with the supply point when the supplier or the occupier changes so it is the stable key for the register.

2. Use the clause to reach the supplier as well as the tenant

Where a lease carries a data sharing clause like the BBP model [15] ask the tenant for the letter of authority it provides for. Then request half hourly data from the supplier or the data collector directly. The article on automating utility data collection covers what a repeatable feed needs.

3. Record the split rule as a versioned method

Where only the incoming meter is measured state the rule and the occupancy assumption with the figure. Where a landlord meter or a meter logged by the BMS records the common parts and central plant deduct that reading before the split is applied. Finding 3 shows one assumption can move the answer by 25%.

4. Label every figure with its data quality grade

Report how much of Category 13 is measured and how much is allocated or estimated. The European Public Real Estate Association (EPRA) asks companies to disclose the proportion of landlord obtained utility data that is estimated [24]. Tenant data deserves the same disclosure.

How Sustainify AI supports tenant energy data collection

A route for tenant data. Unlimited tenant accounts are included in the landlord subscription. Occupiers and suppliers submit against a building and a period into a review queue through Scope 3 data collection.

Files as well as feeds. Carbon data integration takes CSV and Excel through a validating importer. The source file is kept and the person who uploaded it is named against every row so the audit trail starts at import.

The split as a rule. Tenant carbon allocation works by lease or area or consumption with time bound rules. Recharge schedules are locked and versioned per reporting period.

Calculation lineage. Carbon data lineage traces each figure back to the meter reading and source behind it so its data quality grade can be shown to an assurance provider.

The platform supports a submission and does not replace an assurance provider. No estate is ever fully connected and the platform is built on that assumption.

Limitations

The company comparison is small. Six listed landlords define coverage in different ways. British Land reports whole building coverage. The other three figures are occupier or customer measures. LondonMetric counts coverage on assets owned and includes tenant data due after the reporting period. The link between collection route and coverage is an observation and not a measured effect.

The meter statistics cover smaller non domestic sites in the rollout. They count electricity and gas meters together and the bulletin text gives no split by fuel. They do not show how many tenant demises have their own MPAN or how many tenants buy their own supply. Small supplier figures run to December 2025.

Settlement is not access. MHHS changes how suppliers are settled. It gives a landlord no right to tenant data and it does not replace a meter that cannot record half hourly data.

The lease evidence is narrow. The lease analysis covers 26 leases and its sample was skewed toward green buildings.

The studies are not all about leased offices. Hong and colleagues studied schools. Menezes and colleagues studied one office. The performance gap review compares design predictions with measured use rather than benchmarks with measured use. They show the size of error estimation can carry and not the error in any particular portfolio.

The worked example is invented. It shows the effect of one assumption and not a typical result.

What not to conclude. Allocated and estimated figures are not wrong. The GHG Protocol provides for both [6]. The point is to label them.

References

[1] Department for Energy Security and Net Zero. Smart Meters in Great Britain Quarterly Update June 2026 Statistical Bulletin. 27 August 2026. Sections 2 and 2.3. Accessed 6 October 2026.

[2] Elexon. Market wide Half Hourly Settlement. Page last modified 7 September 2026. Accessed 6 October 2026.

[3] MHHS Programme. Key Programme Milestones. Milestone M15. Accessed 6 October 2026.

[4] The British Land Company PLC. Sustainability Progress Report 2026. Page 12. Accessed 6 October 2026.

[5] Hammerson plc. ESG Report 2025. Page 17. Accessed 6 October 2026.

[6] GHG Protocol. Technical Guidance for Calculating Scope 3 Emissions Category 8 Upstream Leased Assets. 2013. Pages 94 to 101 and Formula 8.2. Accessed 6 October 2026.

[7] van Dronkelaar C et al. A review of the energy performance gap and its underlying causes in non domestic buildings. Frontiers in Mechanical Engineering. Volume 1 article 17. 2016. Accessed 6 October 2026.

[8] Meier A and Cautley D. Practical limits to the use of non intrusive load monitoring in commercial buildings. Energy and Buildings. Volume 251 article 111308. 2021. Accessed 6 October 2026.

[9] LondonMetric Property plc. Responsible Business and ESG Report 2026. Pages 2 3 25 and 26. Accessed 6 October 2026.

[10] Financial Conduct Authority. PS26/19 Aligning listed issuers sustainability disclosures with international standards. 30 September 2026. Paragraphs 3.12 to 3.14 and 3.23. Accessed 6 October 2026.

[11] Department for Business and Trade. UK Sustainability Reporting Standard S2 Climate related Disclosures. 25 February 2026. Paragraphs B47 and B48. Accessed 6 October 2026.

[12] GHG Protocol. Technical Guidance for Calculating Scope 3 Emissions Category 13 Downstream Leased Assets. 2013. Pages 128 to 129. Accessed 6 October 2026.

[13] Department for Energy Security and Net Zero. Smart Meters Statistics Methodology Note. 27 August 2026. Accessed 6 October 2026.

[14] Janda KB et al. The evolution of green leases towards inter organizational environmental governance. Building Research and Information. Volume 44 issues 5 and 6 pages 660 to 674. 2016. Accessed 6 October 2026.

[15] Better Buildings Partnership. Green Lease Toolkit Data Sharing Clause. Last updated 11 March 2026. Accessed 6 October 2026.

[16] SEGRO plc. Annual Report and Accounts 2025. Pages 23 and 27. Accessed 6 October 2026.

[17] Derwent London plc. Responsibility Report 2025. Pages 11 and 23 to 24. Accessed 6 October 2026.

[18] Workspace Group plc. 2026 Sustainability Report. Annual Report and Accounts 2026. Pages 82 and 85. Accessed 6 October 2026.

[19] Axon CJ et al. Building communities reducing energy use in tenanted commercial property. Building Research and Information. Volume 40 issue 4 pages 461 to 472. 2012. Accessed 6 October 2026.

[20] Hong SM et al. A comparative study of benchmarking approaches for non domestic buildings Part 1 Top down approach. International Journal of Sustainable Built Environment. Volume 2 issue 2 pages 119 to 130. 2013. Accessed 6 October 2026.

[21] Menezes AC et al. Predicted vs actual energy performance of non domestic buildings using post occupancy evaluation data to reduce the performance gap. Applied Energy. Volume 97 pages 355 to 364. 2012. Accessed 6 October 2026.

[22] Hart GW. Nonintrusive appliance load monitoring. Proceedings of the IEEE. Volume 80 issue 12 pages 1870 to 1891. 1992. Accessed 6 October 2026.

[23] Norford LK and Leeb SB. Non intrusive electrical load monitoring in commercial buildings based on steady state and transient load detection algorithms. Energy and Buildings. Volume 24 issue 1 pages 51 to 64. 1996. Accessed 6 October 2026.

[24] European Public Real Estate Association. EPRA Sustainability Best Practices Recommendations Guidelines Fourth Edition. April 2024. Section 3.4 on page 9 which summarises Section 7.3. Accessed 6 October 2026.