Turning a Year of Meter Reads and a Materials Schedule Into Two Auditable Numbers

Two boxes on the portal, and the folder answers neither

The portal closes at noon on Friday and the sustainability section runs to four questions. One of them wants your most recent full year of electricity restated as tonnes of carbon dioxide equivalent. Another wants the recycled content of the materials you are proposing, as a percentage of materials value. The folder someone has already assembled holds fourteen electricity invoices in three formats, a spend figure lifted out of the accounts system in pounds rather than kilowatt-hours, and a stack of product datasheets whose sustainability paragraph contains the phrase "contains recycled material" and no number at all. Nothing in it drops into either box.

The two questions sit on one page and they are not one question. The electricity figure is an energy accounting exercise: its rules come from the GHG Protocol Corporate Accounting and Reporting Standard, it depends on a boundary you choose and declare, and its arithmetic is a multiplication by a factor with a publication year attached. The recycled content figure is a materials accounting exercise: its definitions come from ISO 14021, its weighting convention comes from the rating system the project is registered under, and the number it produces is a ratio of money even though the content percentages inside it are measured by mass. Running them as one job is how a submission ends up with two figures that are each individually defensible and jointly incoherent.

Both are answerable in an afternoon if the evidence is gathered in the right order and unanswerable for a fortnight if it is not. The ordering rule is the same for both: settle the boundary before touching a number, because everything downstream of a boundary has to be recalculated when the boundary moves, and boundaries always move late.

Which electricity is yours before it is anybody's carbon

The Corporate Standard offers three consolidation approaches — equity share, financial control and operational control — and the answer changes depending on which one you are working to. For a contractor, operational control is usually the least argued-with, because it maps onto the sites you actually run rather than onto a shareholding structure nobody in the bid team can explain. Pick one, apply it to every supply, and name it in the answer. A submission that does not say which basis it used is asking the assessor to guess, and the assessor will guess the unflattering one.

On that basis, in come the temporary builders' supplies taken out in your own name for the duration of a job, the yard, the workshop, and the offices where you hold the meter. Out go the client's supply on a fit-out in a live building — their power, their inventory, your scope 3 — and the subcontractor's own compound supply, which belongs to them for the same reason. None of this is a judgement about who caused the emissions. It is a rule about who reports them, and its purpose is that every kilowatt-hour on a project is counted once by somebody rather than twice by two people or by nobody.

Diesel is the trap. Fuel burned in a generator or a lighting tower you operate is scope 1, not scope 2, and on a year weighted towards groundworks and remote sites it can comfortably exceed everything that came through a meter. A tender asking only for scope 2 therefore hands a flattering answer to whichever bidder ran on generators, which is precisely why the better questionnaires ask for scope 1 and scope 2 together with an intensity ratio underneath. If yours does not, the useful thing is to volunteer the scope 1 figure anyway rather than to win a comparison you know is not a comparison.

Leased and serviced space is the other one worth settling early. Where you rent a floor, the landlord holds the supply and the electricity arrives as a line in a service charge. Whether that consumption is scope 2 or scope 3 depends on the consolidation approach and on whether the leased asset is treated as operated. The Corporate Standard sets out the choice, the Scope 2 Guidance covers how purchased electricity is then accounted for, and the Corporate Value Chain (Scope 3) Standard is where it lands if it is not scope 2. Decide once, write the decision at the top of the working sheet, and resist revisiting it site by site because one treatment produces a smaller total.

Twelve months that are actually twelve months

Consumption, not spend. The figure sitting in the accounts system is money, and money contains standing charges, capacity charges, levies and a unit rate that moved twice during the year. Divide it by anything and you get a fiction. Kilowatt-hours come off the invoices themselves or out of the supplier's portal, and on half-hourly metered supplies they come as an export with the real shape of the year in them.

Then the parts that quietly go wrong. Estimated reads are marked as estimates on the bill, and a run of them followed by a catch-up read dumps a year of drift into a single month, which is fine for a total and disastrous for anything you go on to say about a site. A meter exchange resets the register, so differencing across the change reads as either a year of nothing or a year of everything. Billing periods almost never align to a reporting year, and apportioning the ones that straddle it by days is both correct and something to write down, because it is the first thing that will not reconcile when somebody else repeats your arithmetic.

1. List every supply that was live for any part of the reporting year, by meter point reference, including the ones that opened in March and closed in September.
2. Pull kilowatt-hours rather than pounds for each of them, from invoices or half-hourly exports rather than from the ledger.
3. Flag every estimated read, and replace it wherever actual reads either side let you.
4. Apportion any billing period crossing the start or end of the reporting year by days, and record the apportionment alongside the figure.
5. Stitch the registers together across any meter exchange instead of differencing through it.
6. Add the total back up and reconcile it against the invoices before a factor is applied to anything.

Set the projection to zero unless somebody asked for one

The same calculator will accumulate the figure forward and let you assume the grid decarbonises while it does. A prequalification questionnaire asking for last year's emissions wants none of that. Set the projection period to a single year, set the annual decarbonisation to zero, and what comes out is the figure that belongs in the box: this consumption, at the factor published for that year, and nothing forecast.

The projection earns its keep in a different document. In a whole-life carbon assessment, operational energy is module B6 under EN 15978, and the reporting conventions come from the RICS Whole Life Carbon Assessment for the Built Environment. There, a sixty-year accumulation at a static grid and the same accumulation at an assumed annual fall are two materially different arguments about the same building, and the honest presentation shows both side by side. An assumed decarbonisation is a forecast about policy and investment over decades, not a measurement, and once it is buried inside a single total nobody can see how much of the saving was assumed rather than achieved.

Most questionnaires also want a ratio underneath the absolute, because they are comparing firms of different sizes; under the Streamlined Energy and Carbon Reporting requirements introduced by The Companies (Directors' Report) and Limited Liability Partnerships (Energy and Carbon Report) Regulations 2018, at least one intensity ratio is a requirement rather than a courtesy. Name the denominator you used. Per pound of turnover and per square metre of floor area rank a set of contractors in different orders, and a ratio without its denominator stated is not a ratio.

Recycled content is a defined term, and reuse is not it

ISO 14021 is where the words come from, and it is worth reading rather than paraphrasing. Recycled material is material reprocessed from recovered material. Post-consumer material has finished a life with a household or a commercial user and been recovered from it. Pre-consumer material is diverted from the waste stream during manufacture — and the definition expressly excludes rework, regrind and scrap generated in a process and capable of being reclaimed within that same process. That exclusion is the one suppliers step over most often, because a plant's own offcuts going back into its own hopper look like recycling on a factory tour and represent no diversion whatsoever.

Reuse is a separate category and putting it in this line is a category error a reviewer spots on sight. A salvaged brick relaid in the facade, a steel frame taken down and re-erected, a door set lifted out of one job and hung in another: none of it has been reprocessed, so none of it is recycled content. Rating systems have historically credited reuse under its own heading for exactly that reason, which also means the evidence for it is filed separately and cannot be quietly merged into the recycled content schedule to lift a percentage.

The other thing to be clear about before any arithmetic starts is that the content is declared by mass and the credit is settled in money. A manufacturer states the recycled fraction of a product by weight; that fraction is then applied to what the product cost. Two materials with identical recycled fractions therefore contribute in proportion to their price, which is why the frame, the reinforcement and the concrete decide the answer on almost every job and the ironmongery never does — and why the effort of chasing declarations is worth spending from the top of the cost schedule downwards.

As for what counts as evidence: a product-specific environmental product declaration to EN 15804 and ISO 14025, with ISO 21930 giving the construction-sector core rules, is the strongest thing you can put in front of a reviewer. Next is a dated manufacturer's declaration naming the product and separating post-consumer from pre-consumer content. A marketing line with no figure and no basis is not evidence and should not be entered. For reinforcement the mill route settles it — electric-arc-furnace and blast-furnace bar are different answers — and that is written in the mill's declaration rather than in the chemistry on a test certificate. For aggregate, BS 8500-2 with BS EN 12620 sets out where recycled concrete aggregate is permitted in the UK, and ASTM C33/C33M is the specification the same argument is had against in North America. Where a material's status is genuinely contested — flue-gas desulphurisation gypsum in plasterboard is the standing example, being a by-product of another industry rather than material recovered from a waste stream — the ISO 14021 definitions are what govern, and the defensible move is to state the treatment you applied rather than to adopt whichever reading helps.

Evidence a reviewer can follow without ringing you

Two packs, assembled as the work goes rather than in the last week. The electricity pack holds the list of supplies with their meter point references, the invoices or half-hourly exports behind each, a note of every estimated read replaced and every period apportioned, the factor set named with its publication year, and one sheet that recomputes the total from those inputs rather than restating it. The materials pack holds, per line, the invoice or the supplier's material and labour split, the declaration giving post-consumer and pre-consumer percentages by mass with a date and a product reference, and that line's contribution — plus the single written statement of what the denominator includes and excludes.

The test is not whether your number is right. It is whether somebody who has never seen the job can start from the evidence and arrive at your number without asking you a question. A defensible figure with a stated boundary and a sheet that recalculates beats a better figure that lives only in the head of whoever left in March, and on a framework submission it beats it every year for four years.

1. Name the consolidation approach and the reporting period, once, at the top of both packs.
2. Keep the source data — invoices, exports, declarations — rather than a transcription of it into a spreadsheet.
3. Cite the factor set by title and publication year on the sheet where it is used, not in a covering email.
4. Record every default applied, particularly the 45% material assumption, against the specific lines it was applied to.
5. List the material lines still awaiting a declaration and the share of the schedule they represent, instead of omitting them silently.

The two numbers do not add up to a score

Nothing combines them, and something on the client's side of the table will eventually ask you to try. Operational carbon is a rate: emissions per year of use, at a factor that will most likely be lower next year whatever you do. Recycled content is a share of one project's materials spend at one moment in its procurement. Averaging them, weighting them into an index, or presenting a single sustainability percentage is not a simplification but a fabrication, and the tenders worth winning are read by people who have watched it attempted.

Neither figure is embodied carbon either, which is the third question these pages tend to ask and a genuinely separate exercise, governed by EN 15804 at product level and EN 15978 at building level. High recycled content does correlate with a lower declared product GWP, most visibly in steel, but a correlation is not a declaration and it does not license estimating one from the other.

The sentence that goes in the box

Write each answer so its boundary travels with it. For electricity: the reporting period, the total kilowatt-hours, how many supplies stand behind that total, the factor set by name and publication year, and the location-based and market-based results each labelled as such. For materials: the recycled content value as a percentage, the total materials cost it was taken against, the weighting convention stated in words rather than assumed, and the exclusions. Neither sentence is long. Both are reproducible, which is the property being assessed.

Then say what is not in it. The scope 1 fuel, if you were asked only for scope 2. The transmission and distribution losses, if they are reported elsewhere. The sites that entered the boundary part-way through the year. The material lines whose declarations had not arrived, with the proportion of the schedule they cover. Every one of those is a question a reviewer would otherwise raise on a clarification round, and answering it unprompted is the cheapest credibility available on a bid.

A figure nobody can reproduce is worth less than a slightly worse one with a method under it. That holds for the carbon number and the recycled content number equally, and it is very nearly the only thing the two of them have in common.