Separate Factory Energy Decisions From Reporting Scope

02A finance director may need to establish reporting obligations.

03These questions can share data without sharing the same boundary or decision process.

04The current ESOS phase-four guidance defines a large undertaking by at least 250 employees, or both turnover above GBP 44 million and a balance-sheet total above GBP 38 million.[

051] Apply the full guidance, including group and qualification details.

06Do not infer exemption because one site employs fewer people.

07Equally, do not tell every small manufacturer that installing an energy dashboard is a statutory requirement.

08Establish the actual scheme position with the responsible adviser.

09SECR is a separate reporting framework covered by government guidance; the recent Regulatory Policy Committee review describes its energy and emissions reporting function.[

102][3] Keep that assessment distinct from the ESOS determination.

02A spreadsheet can support the work if its definitions and sources are clear.

03Software sophistication does not compensate for an unknown meter boundary.

04Identify whether a reading covers the whole site, a shared building or a particular production process.

05Record units, meter identifiers and the period represented.

06Ask who investigates gaps or estimated readings.

07Keep source readings unchanged and calculate derived measures separately.

08If an invoice is corrected or a meter replaced, preserve the relationship between the original evidence and the revised calculation.

09The tradeoff is measurement detail.

10Submetering every machine may be expensive to maintain, while a single site total may hide the issue.

11Start with the operational question and choose the smallest useful measurement boundary.

02It does not by itself demonstrate better energy efficiency.

03Separate consumption, cost and output when presenting the result.

04Record production context such as operating hours, product mix and planned shutdowns.

05A month producing heavier or more complex work may not be directly comparable with a quieter month.

06Use intensity measures carefully.

07Energy per unit can help for consistent output, but it may mislead when units vary substantially.

08Explain the denominator and the conditions under which comparisons are meaningful.

09AI can identify unusual periods and draft investigation questions.

10Require the operations owner to assess explanations against site knowledge before the business reports a saving or changes a production practice.

02It records 120 kWh on each night, giving 10 times 120, or 1,200 kWh.

03After a reviewed operational change, a further ten comparable nights average 95 kWh, totalling 950 kWh.

04The observed difference is 1,200 minus 950, or 250 kWh across those periods.

05At an assumed illustrative energy rate of GBP 0.20 per kWh, 250 kWh corresponds to GBP 50 of energy charges.

06This excludes other charges and does not use a claim about current tariffs.

07Before describing the difference as a saving caused by the change, examine weather, occupied hours, production spillover and measurement consistency.

08Do not annualise ten nights automatically or convert the difference into carbon emissions without an appropriate methodology and factors.

02A reporting process needs to decide whether and how those estimates are acceptable for its methodology.

03Label estimates and assumptions clearly.

04Keep the original evidence accessible to the person preparing any applicable report or assessment.

05A polished AI narrative cannot replace the calculation trail.

06Use approved conversion factors and accounting boundaries where emissions reporting is required.

07Record the version and source through the reporting process, rather than letting a model recall a factor from general knowledge.

08The failure case is one total reused everywhere: purchasing treats it as cost, operations calls it efficiency and marketing calls it a carbon reduction.

09These are different claims, requiring different calculations and evidence.

02Avoid making an energy-management platform the prerequisite for understanding one clear anomaly.

03Agree the comparison period, relevant context and who will review the result.

04Include a plan for missing data and unexpected changes.

05This makes the analysis interpretable even when the first hypothesis is wrong.

06Keep the compliance owner informed about useful evidence without implying that this exercise fulfils ESOS or SECR requirements.

07The applicable formal work may need additional scope, methodology and qualified involvement.

08The next deliverable is a traceable baseline and one reviewed operating decision.

09That creates practical value while preserving the separate entity, group and reporting judgements required for the business's actual UK obligations.