Most contractors start measuring emissions with fuel and electricity, because the data is easy: fuel cards and utility bills. That's the right first step, and our guide to a minimum viable carbon answer covers scopes 1 and 2 in detail. But for most construction businesses, the bulk of the footprint sits in scope 3, and the largest slice is category 1: purchased goods and services. Concrete, steel, aggregates, asphalt, timber, plant hire and subcontracted work all land there.
Scope 3 looks intimidating because it's about things you don't control. In practice, you can get a useful first estimate from your purchase ledger in a few days, then sharpen it where it matters. This guide covers the categories that count, the three calculation methods, a worked example, hotspots, what to ask suppliers, honest tender answers and the reduction levers that move the number.
The scope 3 categories that matter for a contractor
The GHG Protocol's scope 3 standard splits value chain emissions into 15 categories. Most of them are about selling products, so a contractor can usually set them aside. The ones that typically matter:
| Category | What it covers for a contractor | Typical size | Where the data is |
|---|---|---|---|
| 1. Purchased goods and services | Materials (concrete, steel, aggregates, asphalt, timber, pipes, precast), plant hire, subcontractors, consultants, office and IT | Usually the largest by far | Purchase ledger, delivery tickets, supplier data |
| 2. Capital goods | Plant, vehicles and equipment you buy outright | Lumpy: large in years you buy kit | Fixed asset register |
| 3. Fuel and energy related | Emissions from producing and delivering the fuel and electricity you use (well-to-tank, grid losses) | Small, easy to add | Your scope 1 and 2 data, with the matching factors |
| 4. Upstream transport | Delivery of what you buy to your sites, and haulage you pay for, in vehicles you don't own | Moderate; high for heavy, bulky materials | Haulage invoices, delivery tickets, distances |
| 5. Waste generated in operations | Treatment and disposal of site and office waste, including muck-away | Moderate | Waste transfer notes, waste contractor reports |
| 6. Business travel | Flights, rail, hotels, staff cars on mileage | Small | Expense claims |
| 7. Employee commuting | Staff travel to sites and offices | Small to moderate | A short staff survey |
Two rules help. Capital goods count in full in the year you buy them, so a year of new excavators jumps. And hire plant fuelled by the hire company sits in your scope 3, not scope 1; set the boundary once, as the minimum viable answer guide explains, and stick to it.
Three ways to calculate category 1
The GHG Protocol's guidance describes several methods. Three matter for a contractor, and you'll usually use all three at once for different parts of your spend.
| Method | How it works | Accuracy | Effort | Best for |
|---|---|---|---|---|
| Spend-based | Money spent × an emission factor per pound, dollar or euro for that type of product | Low | Low | A first pass across everything; services; long tail of small purchases |
| Average-data | Quantity bought (tonnes, m³, metres) × an average emission factor for that material | Medium | Medium | Your top material categories, before supplier data is available |
| Supplier-specific | The supplier's own figure for the product you bought, usually from an EPD | Highest | Higher | The few materials and suppliers that dominate your footprint |
Spend-based: fast, rough, and blind to good choices
Spend-based factors come from economic input-output models of the average emissions per unit of money spent in an industry. Sources include the UK government's supply chain emission factors by industry (published alongside its carbon footprint statistics), the US EPA's supply chain emission factors for US industries and commodities, and multi-regional databases such as Exiobase. Use one source consistently.
Two cautions. Factors are tied to a price year, so adjust for inflation if your spend is newer than the factors, or you'll overstate. And spend-based figures move with price, not carbon. If you pay more for a lower-carbon concrete mix, a spend-based figure goes up. That's why you shouldn't stay on spend-based for your biggest categories: it can't show the reductions you make.
Average-data: tonnes and cubic metres
You usually already know material quantities from delivery tickets, purchase orders and bills of quantities. Multiply by an average factor for the material.
A common source in UK construction is the ICE database (Inventory of Carbon and Energy), originally developed at the University of Bath and now maintained by Circular Ecology. It gives embodied carbon factors for common construction materials, typically "cradle to gate": everything up to the product leaving the factory. The UK government's annual conversion factors also include some material factors. Whichever you use, note the version and the boundary.
Supplier-specific: EPDs
An Environmental Product Declaration (EPD) is a verified, standardised statement of a product's environmental impacts, including its global warming potential. In Europe and the UK, construction product EPDs follow EN 15804; in North America, ISO 21930 is common. Both build on ISO 14025.
To use one, check:
- Declared unit. The figure is per something: per m³ of concrete, per tonne of steel, per m² of board. Match it to your quantities.
- Life cycle modules. A1 to A3 is the product stage (raw materials, transport to the factory, manufacturing). A4 is transport to site, A5 is construction. For category 1, you normally use A1 to A3, and count delivery in category 4.
- Product-specific or industry-average. One manufacturer's product, or a sector average. Both beat a generic factor; say which you used.
- Validity. EPDs are typically valid for five years.
- Timber and biogenic carbon. Timber EPDs may show a negative A1 to A3 figure because they count carbon stored in the wood. Don't silently net that off against everything else; report fossil emissions and any stored carbon separately and say so.
EPDs are published through programme operators such as BRE in the UK, the International EPD System and IBU, and databases such as EC3 aggregate many of them.
A worked example
Here's an illustrative 45-person civils and groundworks contractor with £9 million turnover. Its scope 1 and 2 footprint, mostly diesel, is around 250 tCO₂e. All factors below are illustrative and rounded, chosen to show the method; use current published factors and your suppliers' data for real figures.
Step 1: a spend-based first pass
Group a year of the purchase ledger:
| Ledger category | Spend | Illustrative factor (kg CO₂e per £) | Estimate |
|---|---|---|---|
| Ready-mix concrete | £1,000,000 | 1.6 | 1,600 tCO₂e |
| Steel (rebar, mesh, fabrications) | £300,000 | 1.4 | 420 tCO₂e |
| Aggregates and fill | £450,000 | 0.5 | 225 tCO₂e |
| Asphalt and surfacing | £250,000 | 0.9 | 225 tCO₂e |
| Pipes, ducts and drainage products | £350,000 | 0.7 | 245 tCO₂e |
| Timber and formwork | £100,000 | 0.4 | 40 tCO₂e |
| Plant hire (with operators and fuel) | £700,000 | 0.4 | 280 tCO₂e |
| Subcontractors | £1,400,000 | 0.25 | 350 tCO₂e |
| Haulage (category 4) | £300,000 | 0.6 | 180 tCO₂e |
| Waste and muck-away (category 5) | £250,000 | 0.5 | 125 tCO₂e |
| Professional services, IT, office | £200,000 | 0.15 | 30 tCO₂e |
| Total | £5,300,000 | 3,720 tCO₂e |
Even rough, this says a lot. Scope 3 is roughly fifteen times scope 1 and 2. Concrete alone is over 40% of it, and concrete, subcontractors and steel together are about two thirds.
Step 2: refine the biggest categories
Delivery tickets show 7,000 m³ of concrete and 320 tonnes of rebar and mesh. The contractor applies an average factor, then asks its two main suppliers for EPDs:
| Material | Quantity | Average-data (illustrative) | Supplier-specific (illustrative) |
|---|---|---|---|
| Concrete | 7,000 m³ | 280 kg CO₂e/m³ → 1,960 t | 4,500 m³ at 220 + 2,500 m³ at 180 → 1,440 t |
| Rebar and mesh | 320 t | 1.9 t CO₂e/t → 608 t | 0.75 t CO₂e/t (electric arc furnace, high scrap content) → 240 t |
The spend-based concrete figure (1,600 t) sat between the generic average (1,960 t) and the supplier figure (1,440 t). For steel, the generic factor was more than double the supplier's, because production route matters enormously: electric arc furnace steel made mostly from scrap usually carries far less carbon than blast furnace steel. Average factors can be wrong in either direction; only supplier data tells you about what you actually bought.
With those rows updated, scope 3 becomes about 3,380 tCO₂e. More importantly, the number will now fall if the contractor switches to lower-carbon mixes, and it can show that it did.
Finding your hotspots
A hotspot is a category that is both large and within your influence. Rank categories by emissions, then ask three questions:
- How big is it? Your top three to five categories will usually cover most of the total.
- Can we influence it? You choose your concrete supplier and can propose mixes; you have less say over a client-specified product or a nominated subcontractor.
- How good is the data? A big category estimated by spend is a priority for better data, even before you try to reduce it.
The GHG Protocol suggests judging data quality on technology, time period, geography, completeness and reliability. A high, medium or low rating per category is enough to show where next year's effort goes.
Do this by project as well as by year: one concrete-heavy job can dominate a year, and clients increasingly ask for project figures.
What to ask your suppliers
Start with the five to ten suppliers behind your top categories, and be specific:
- Do you have an EPD for the products we buy from you? Product-specific or industry-average? Which standard, which modules, and when does it expire?
- For concrete: what's the embodied carbon per m³ of the mixes we order, and what lower-carbon alternatives can you supply at the same strength class (for example with cement replacements such as GGBS or fly ash)? What are the implications for curing time and early strength?
- For steel: which production route and what recycled content? Can you provide mill or product EPDs?
- For aggregates and asphalt: recycled content, distance from quarry or plant to our sites, and whether warm-mix or high recycled-content asphalt is available.
- For hauliers and waste contractors: distances and fuel data; for waste, tonnes by stream and destination.
Explain why: your clients are asking you. Store every EPD and response in one place with its expiry date, as you would insurance certificates, so next year starts from what you already hold.
Answering tender questions honestly
Scope 3 questions in tenders tend to come in a few forms: "Do you measure scope 3?", "What are your largest sources of emissions?", "Provide embodied carbon figures for the materials on this project", and "How will you reduce embodied carbon on this contract?"
For UK public sector bidders, note that the Carbon Reduction Plan format for major central government contracts (introduced as PPN 06/21) has historically asked for a specified subset of scope 3 categories: upstream transport, waste, business travel, employee commuting and downstream transport, rather than purchased goods. Check the current guidance and the tender wording for exactly what's required. Many clients and frameworks ask about materials anyway, because that's where their own scope 3 sits.
A strong answer has the same parts whatever the wording:
Method. We estimate scope 3 annually using the GHG Protocol. For [year], [X]% of our category 1 emissions were calculated from supplier EPDs, [Y]% from material quantities using ICE factors, and the remainder from spend-based factors from [source].
Results. Our estimated scope 3 emissions were [3,380] tCO₂e. Our largest sources are concrete ([43]%), subcontracted work and steel.
Actions. We now ask for EPDs from all concrete and steel suppliers, have trialled [mix] on [project], reducing concrete embodied carbon on that job by an estimated [Z]%, and [second action].
On this contract. We will propose lower-carbon mix options for the designer's approval, report material quantities and EPD-based embodied carbon [monthly / at completion], and report waste by stream.
Limitations and next steps. Our subcontractor emissions are spend-based estimates. In [year] we will request data from our ten largest subcontractors.
Being upfront about gaps isn't a weakness. A precise-looking number with no method behind it invites hard follow-up questions. If you reuse this answer across bids, keep it in your answer bank and update the figures each year; the approach in answering PQQs without starting from a blank page applies directly.
Reduction levers that actually move the number
Focus on your hotspots. For most contractors that means materials first.
Concrete
- Cement replacement. Mixes using GGBS, fly ash or other supplementary materials can cut embodied carbon substantially compared with a straight CEM I mix. Availability varies by region, so ask your supplier what they can reliably deliver.
- Specify by performance. Where the design allows, specifying strength, durability and exposure class rather than cement content gives the supplier room to optimise.
- Allow longer strength gain. Higher replacement mixes often gain strength more slowly; accepting 56-day strength where the programme allows can unlock them.
- Use less. Optimised slabs and foundations and accurate ordering reduce volume, the most reliable reduction of all.
Steel
- Production route and recycled content. Ask for electric arc furnace steel with high scrap content where it meets the specification.
- Reuse reclaimed sections where testing and the engineer allow, and rationalise over-specified sections and reinforcement.
Aggregates, asphalt and earthworks
- Recycled and secondary aggregates where the specification permits.
- Reuse excavated material on site. A better cut and fill balance cuts imported aggregate, haulage and muck-away: categories 1, 4 and 5 at once.
- Warm-mix asphalt and recycled asphalt reduce production energy and virgin material.
- Source locally. For heavy materials such as aggregates, transport can be a large share of the total.
Waste and transport
- Segregate waste on site, using waste contractors who report by stream.
- Order accurately, use take-back schemes for offcuts and packaging, and consolidate deliveries.
Design and value engineering with the client
The biggest savings are usually made at design stage, and often aren't yours alone to make. Propose alternatives early, with carbon and cost side by side. Never substitute a specified material or change a design without the designer's and client's approval. Structural, durability and warranty consequences come first. Where clients work to PAS 2080, the carbon management standard for buildings and infrastructure, they'll expect suppliers to bring this kind of challenge as part of the job.
Common mistakes
- Staying on spend-based for your biggest categories. It's fine as a start, but it can't show reductions and penalises paying more for better products.
- Mixing boundaries, such as counting delivery in both category 1 and category 4.
- Netting off timber's stored carbon without saying so.
- Using expired or mismatched EPDs, or applying a per-tonne figure to cubic metres.
- Quoting a total with no method. State the year, categories, methods and share of supplier data.
- Claiming reductions you can't evidence. Keep the delivery tickets and mix data.
Where software helps
Most of this work is gathering and tracking: grouping ledger lines, matching delivery tickets to quantities, storing EPDs with expiry dates, and drafting tender answers from checked figures. Software can take that on. Which boundary to draw, which factors to trust and what you commit to on a contract stay your call.
Start here
Export last year's purchase ledger this week and group it into a dozen categories. Apply one published set of spend-based factors and rank the results. You'll almost certainly find that two or three materials dominate. Then pick the top one, usually concrete, pull the quantities from your delivery tickets and ask that supplier for EPDs for the mixes you buy. That one step turns your biggest number from a guess into a figure you can defend in a tender and reduce over time.
Measure roughly everywhere, then precisely where it counts.