The Carbon Contingency: Should Projects Carry an Emissions Allowance?

Mar 11

16 min read

A project can carry a sizeable cost contingency, a schedule reserve, and a risk allowance for ground conditions, weather, fabrication delays, marine access, and commissioning trouble. Yet the same project may present its greenhouse gas estimate as a neat number, often to three significant figures, long before the design is mature enough to justify that precision.

That gap is becoming harder to defend.

Major projects now face closer scrutiny from boards, lenders, regulators, joint venture partners, insurers, customers, and local communities. Emissions are no longer a side calculation in an environmental appendix. They can affect permits, financing conditions, procurement choices, product competitiveness, carbon pricing exposure, abandonment liabilities, and long-term asset strategy.

This raises a practical project-control question: should major projects carry a quantified carbon contingency, an allowance for uncertainty in estimated emissions, much like cost or schedule contingency?

The answer is not a simple yes. A carbon contingency can improve transparency and decision-making if it is governed well. It can also become a soft hiding place for weak estimating, late design change, or a lack of emissions-reduction discipline. Used badly, it can make carbon management look mature while doing little to reduce real emissions.

Used properly, it can become one of the most useful tools in a project director’s ESG kit.

Why project carbon estimates are uncertain early on

Carbon estimates are only as strong as the design, quantities, assumptions, emission factors, and boundaries behind them. At concept, feasibility, and FEED, those inputs are rarely settled.

In an oil and gas development, the team may not yet know the final module weights, well count, compression duty, flare philosophy, pipeline route, vessel spread, or power source. In a mining project, the haulage strategy, plant throughput, grid connection timing, camp size, explosives use, and tailings design may still be moving. In a transport infrastructure program, alignments, pavement designs, bridge spans, soil treatment, tunneling methods, and possession windows can change materially between option selection and detailed design.

That matters because project emissions come from many places, not one line item.

They can include:

Embodied carbon in steel, cement, aluminum, plastics, insulation, coatings, cables, pipe, equipment, and prefabricated modules.
Construction emissions from earthworks, piling, cranes, concrete batching, welding, site plant, waste, accommodation, and temporary facilities.
Logistics emissions from marine transport, heavy lift vessels, road haulage, aviation, laydown yards, and expedited freight.
Temporary power from diesel generation, gas turbines, grid imports, battery systems, or hybrid arrangements.
Supplier emissions data that may be based on generic factors, product-specific declarations, or commercial estimates.
Commissioning emissions from flushing, pressure testing, drying, steam blows, flaring, venting, energization, and reliability runs.
Operational emissions over the asset life, including fuel use, fugitive emissions, purchased electricity, process emissions, losses, and maintenance activity.
Future design changes driven by safety, constructability, climate resilience, regulation, debottlenecking, operability, or owner requirements.

International frameworks already recognize, in different ways, that greenhouse gas accounting depends on boundaries, data quality, assumptions, and uncertainty. The GHG Protocol separates scopes and categories. ISO 14064 and ISO 14067 address greenhouse gas quantification and product carbon footprints. Life cycle assessment practice under the ISO 14040 family pays close attention to system boundaries and data quality. IFRS S2, issued by the International Sustainability Standards Board, pushes companies to disclose material climate-related risks and opportunities in a way investors can use. None of these frameworks magically removes project uncertainty.

The IPCC has long emphasized that emissions pathways depend on technology choices, infrastructure, energy systems, and demand patterns. The IEA’s work on energy transitions points to the lock-in effect of long-lived assets. Those points are highly relevant to capital projects. A pump selection, plant layout, logistics contract, or grid connection decision made during FEED can shape emissions for decades.

Project professionals understand this pattern from cost and schedule. Early estimates have wide ranges. As scope definition improves, estimating confidence improves. PMI and APM risk management guidance both treat uncertainty as something to identify, assess, allocate, monitor, and control. Carbon estimation should not get a pass just because it sits under ESG rather than cost engineering.

A carbon number produced during concept can be useful. It can compare options, flag hot spots, and guide design choices. But it should not be mistaken for a high-confidence baseline unless the project has earned that confidence.

What a carbon contingency is, and what it is not

A carbon contingency is a quantified allowance for uncertainty in a project’s estimated greenhouse gas emissions. It sits on top of, or alongside, a base emissions estimate to reflect known unknowns and estimating maturity.

A simple expression might look like this:

`Forecast emissions = base estimate + carbon contingency`

That sounds straightforward, but the governance matters. The allowance should not be a casual percentage added at the end. It should reflect the project’s risk profile, design maturity, data confidence, procurement status, construction methods, logistics plan, commissioning philosophy, and operational assumptions.

It may reduce over time as the project matures:

Project phase	Typical emissions uncertainty drivers	What should improve
Concept	Broad technology options, incomplete quantities, generic emission factors, unclear boundaries	Option-level carbon ranges and major hot spots
Feasibility	Early layouts, preliminary equipment lists, immature supply chain data, route uncertainty	Better quantity estimates and clearer lifecycle boundaries
FEED	Defined basis of design, early vendor engagement, draft construction plan, early logistics strategy	More reliable embodied and construction carbon estimate
Detailed design	Final quantities, vendor data, package awards, construction sequencing	Lower contingency and stronger change control
Execution	Actual fuel use, delivery records, supplier declarations, commissioning data	Replacement of estimates with measured or supplier-specific data

This is familiar territory for project teams. Cost contingency is not the approved budget. Schedule contingency is not the same as the project completion date. Risk allowance is not permission to manage badly. In the same way, carbon contingency should represent uncertainty, not ambition.

It must also be separated from nearby concepts.

A carbon contingency is not a carbon offset. Offsets are credits or removals used to compensate for emissions elsewhere, subject to concerns about quality, permanence, additionality, and verification. A contingency is an allowance for uncertainty in the project forecast. It does not neutralize emissions.

It is not an emissions target. A target describes a desired outcome, often linked to corporate climate strategy or asset performance. A contingency describes uncertainty around the forecast.

It is not a carbon budget. A carbon budget sets an allowed emissions envelope for a project, program, asset, or portfolio. A contingency may sit within that envelope or be governed against it.

It is not a planned growth allowance. Some assets include future capacity, tie-ins, debottlenecking space, or phased expansion. Those are planned scope or strategic optionality. Carbon contingency should not hide future growth that is already reasonably foreseeable.

This distinction matters commercially. If a project uses “contingency” to disguise planned capacity expansion, then the baseline is weak. If it uses contingency to reflect genuine uncertainty, the decision record becomes clearer.

Where emissions uncertainty enters the project

The most productive carbon contingency discussions are specific. They do not ask, “How uncertain is carbon?” They ask, “Which emissions sources are uncertain, why, and what will reduce that uncertainty?”

Embodied carbon

Embodied carbon is often one of the largest early blind spots. At concept stage, teams may use benchmark quantities or high-level material take-offs. Steel tonnage, concrete volumes, pipe classes, foundation sizes, cable lengths, insulation thickness, and temporary works can move significantly as design develops.

Supplier data adds another layer. A generic emission factor for structural steel is different from a product-specific environmental product declaration. Cement blends, recycled content, manufacturing energy, transport distance, fabrication methods, and country of origin all affect the estimate.

Peer-reviewed life cycle assessment literature consistently shows that results are sensitive to boundary choices, data quality, allocation methods, and assumptions. That does not make LCA unreliable. It means project teams should be honest about confidence levels and should avoid false precision.

Construction and installation

Construction-phase emissions are often treated as secondary compared with operational emissions. That may be reasonable for some long-life process assets, but not for all projects. In civil infrastructure, marine works, offshore installation, remote mining projects, and renewable energy construction, the build phase can be material.

Uncertainty can come from:

Ground conditions and rework
Weather downtime and standby fuel
Crane and vessel selection
Temporary roads, camps, and laydown areas
Concrete curing, dewatering, and pumping
Welding rates and generator loading
Waste volumes and disposal routes

A marine project provides a simple example. The emissions estimate may assume a certain vessel class, day rate, fuel type, and installation duration. If the vessel spread changes, or weather windows extend the campaign, emissions can move quickly. A cost estimate would normally recognize that risk. A carbon estimate should too.

Logistics and supply chain routes

Capital projects move heavy things across long distances. Modules may come from fabrication yards in different regions. Specialist equipment may travel by sea, road, air, or a mix of all three. Expedited freight can appear late when schedule pressure rises.

The emissions difference between planned sea freight and urgent air freight can be material. So can the difference between a local supplier and a distant supplier if the distant supplier has lower manufacturing emissions but higher transport emissions. There is no universal answer. The point is that logistics assumptions need to be visible.

Procurement teams can help reduce contingency by requiring better supplier data, transport plans, packaging assumptions, and delivery schedules. They can also avoid contract structures that push suppliers to choose lower-cost but higher-emission routes without disclosure.

Temporary power

Temporary power is a common undercount. Site teams may start with a rough estimate for diesel generation, then add lighting towers, camps, workshops, pumps, welding sets, offices, refrigeration, dewatering systems, and backup capacity as reality sets in.

In some regions, grid power is available early. In others, grid connection lags the construction program. Hybrid solutions using batteries, solar, gas, or grid imports may reduce emissions, but they require planning, space, reliability assessment, and commercial alignment.

A carbon contingency can force the temporary power conversation earlier, before diesel becomes the default because every other option has missed the procurement window.

Commissioning and start-up

Commissioning emissions are frequently underestimated because they sit between construction and operations. They can include steam generation, fuel gas use, flaring, venting, nitrogen purging, drying, chemical cleaning, reliability testing, trips, restarts, and off-spec production.

Standards and industry guidance from organizations such as IOGP and the Energy Institute support more consistent reporting of oil and gas emissions sources, including methane and flaring in relevant contexts. The discipline is mature enough to do better than rough assumptions, but commissioning plans are often still immature at FEED.

A quantified uncertainty allowance can make commissioning emissions visible at the stage when design choices, operating procedures, and equipment selection can still reduce them.

Operational emissions and lifecycle forecasting

For many industrial assets, operational emissions dominate lifecycle emissions. They depend on availability, throughput, energy efficiency, fuel choice, reservoir behavior, ore grade, process chemistry, grid carbon intensity, maintenance cycles, methane management, leakage rates, and end-user demand.

Forecasting those emissions is not just an ESG exercise. It connects to asset economics, carbon pricing, product market access, lender expectations, and potential future regulation.

The World Bank and other institutions have documented the global spread of carbon pricing instruments. The IFC Performance Standards and World Bank Environmental, Health, and Safety Guidelines have also influenced lender expectations for environmental and social risk management. Project sponsors do not need perfect policy foresight to see the issue. A high-emitting design carries commercial exposure if carbon constraints tighten.

A carbon contingency should not replace scenario analysis. It can sit beside it. The contingency captures uncertainty in estimates around the selected project case. Scenario analysis tests different future conditions.

How carbon contingency can fit into project controls

The strongest argument for a carbon contingency is not that it creates a new ESG metric. It is that it connects emissions to the normal machinery of project governance.

That machinery already exists. Stage gates, baselines, risk registers, cost reports, procurement strategies, design reviews, management of change, and assurance processes are standard in capital-intensive industries. Carbon management becomes more credible when it uses those controls rather than sitting in a parallel sustainability report.

Carbon budgets and project baselines

A project carbon budget sets the emissions limit or envelope approved by the sponsor. It may cover construction only, lifecycle emissions, or specific categories such as embodied carbon and Scope 1 construction fuel. The boundary must be explicit.

The baseline is the current approved estimate against that budget. If the base estimate is 100 units and the contingency is 20 units, leadership should know whether the approved carbon budget is 100, 120, or something else. This is not a semantic issue. It affects decisions.

If the budget includes contingency, drawdown rules are needed. If the budget excludes contingency, the project needs escalation triggers when forecast emissions exceed the budget. Either model can work, but ambiguity cannot.

Stage gates

At each gate, the project should present both the emissions forecast and its confidence level. A gate paper that shows a single carbon number without uncertainty tells decision-makers less than it should.

A more useful gate review would ask:

What is the base emissions estimate?
What contingency has been applied and why?
Which assumptions drive the widest uncertainty?
Which risks could push emissions beyond the budget?
What evidence would allow contingency to be reduced before the next gate?
Which design decisions are irreversible or costly to change later?

This mirrors cost and schedule practice. It also aligns with the direction of climate-related financial disclosure. Investors and lenders want decision-useful information, not polished sustainability language with weak controls behind it.

Procurement

Procurement is where carbon estimates either improve or decay. If bid documents do not ask for product emissions data, fuel assumptions, logistics routes, fabrication locations, or environmental product declarations where relevant, the project may remain dependent on generic factors until late in execution.

Procurement can reduce carbon contingency by improving evidence. It can include carbon data requirements in supplier packages, define acceptable calculation methods, request product-specific information, and make clear how carbon will be evaluated against cost, schedule, quality, safety, and local content.

This does not mean lowest-carbon always wins. Capital projects involve trade-offs. A supplier with lower embodied carbon may carry schedule risk. A local supplier may reduce logistics emissions but have higher manufacturing emissions. A lower-carbon construction method may add cost or interface complexity. Good governance makes those trade-offs visible.

Change control

Change control is where carbon contingency may deliver its greatest value. Many emissions increases don't come from one major decision. They accumulate through design revisions, late procurement changes, temporary works, rework, construction acceleration, and commissioning extensions.

If carbon is included in management of change, each material change should show its emissions effect. That does not require a full lifecycle assessment for every valve substitution. It does require proportional rules.

A change that increases steel tonnage, diesel generation, flaring duration, vessel days, concrete volume, or operating energy demand should not pass through the system with only cost and schedule impacts recorded.

Risk management

A carbon contingency belongs in the same risk conversation as cost, schedule, safety, quality, and operability. It should link to named risks, not float as a percentage.

For example:

Risk	Carbon effect	Possible control
Grid connection arrives late	More diesel or gas temporary power	Early utility agreement, hybrid power plan, staged energization
Ground conditions require redesign	More concrete, steel, equipment hours, waste	Better site investigation, geotechnical hold points
Vendor data unavailable	Higher uncertainty in embodied carbon	Procurement data requirements, approved emission factor hierarchy
Offshore campaign extends	More vessel fuel and standby emissions	Weather planning, vessel assurance, installation readiness
Commissioning instability	More flaring, fuel use, venting, off-spec operation	Commissioning emissions plan, start-up readiness reviews

This is not theoretical. It is project control applied to emissions.

The commercial case for using a carbon contingency

The case for carbon contingency is strongest where emissions uncertainty can change decisions. If the number will not affect anything, it is just another reporting artifact.

In high-risk capital projects, it can affect several live choices.

During option selection, it can show that one concept has a lower base estimate but a much wider uncertainty range. A gas-fired temporary power plan may look cheaper and simpler than an early grid or hybrid option, but the emissions risk may be higher if construction extends. A modular construction strategy may reduce site emissions but increase marine logistics. Tunneling may reduce surface disruption but increase energy demand.

During sanction, it can give the investment committee a clearer view of climate-related exposure. A project with a narrow carbon range may deserve different treatment from one with the same base estimate but high uncertainty in operational emissions, fugitive methane, or future power supply.

During execution, it can prevent slow drift. If the team draws down carbon contingency every time scope grows, expedition increases, or rework occurs, leadership sees the pattern. The discussion becomes less emotional and more factual.

There is also a disclosure angle. IFRS and ISSB standards are not project management manuals, but they raise expectations for governance, risk management, metrics, and targets on material climate-related matters. Project-level carbon controls can support corporate reporting by improving the quality of internal information.

Regulators are also moving in this direction in many jurisdictions. Environmental impact assessment processes increasingly expect greenhouse gas considerations, especially for energy, infrastructure, transport, mining, and industrial facilities. The exact requirements differ by country and sector, but the trajectory is clear: emissions estimates need a rational basis and a transparent boundary.

In Gulf markets, including the UAE and wider region, this matters for another reason. Large industrial and infrastructure projects often involve international lenders, export credit agencies, national oil companies, global EPC contractors, and multinational suppliers. The project may sit in one jurisdiction, but its finance and supply chain expectations are global. A transparent emissions contingency can help align discussions among owners, contractors, lenders, and suppliers before disputes arise.

The limitations and risks are real

A carbon contingency can improve decision quality. It can also weaken discipline if the organization treats it as spare room.

The first risk is padding. Teams may add a broad allowance because they do not trust the estimate, without identifying the drivers. That is not contingency management. It is concealment.

The second risk is double counting. If lifecycle emissions scenarios already include future throughput variation, and the contingency also adds an allowance for the same thing, the forecast may overstate uncertainty. The same can happen when supplier emission factors already include conservative assumptions.

The third risk is normalizing overruns. A cost contingency does not make overspending acceptable. A carbon contingency should not make emissions increases acceptable. The project still needs reduction plans, design challenges, and accountability.

The fourth risk is weak baselines. If the baseline is reset too often, every increase can look harmless. Baseline updates are sometimes necessary, especially after a sanctioned scope change. But they should require approval, evidence, and a record of what changed.

The fifth risk is confusing uncertainty with poor performance. Some uncertainty is legitimate. Poor estimating, weak supplier engagement, missing construction data, or late commissioning planning should not be protected by the same language.

Good governance makes these distinctions visible.

A useful carbon contingency should be:

Evidence-based
It should tie to quantities, emission factors, supplier data quality, design maturity, and named risks.
Time-bound
It should reduce as definition improves, unless new risks emerge.
Owned
It should have accountable owners across engineering, procurement, construction, commissioning, operations, and sustainability.
Controlled
Drawdown should require approval through change or risk processes.
Reported with the base estimate
Leaders should see both the estimate and the uncertainty, not a single blended number that hides confidence.

Carbon contingency also has cultural limits. If project incentives reward only capital cost and schedule, carbon contingency will not carry much weight. If procurement treats carbon data as optional, uncertainty will stay high. If operations are not involved early, lifecycle emissions forecasts may miss practical realities.

What leaders should ask before approving one

A carbon contingency is only useful if senior decision-makers ask disciplined questions. The aim is not to turn executives into carbon accountants. The aim is to make uncertainty useful for decision-making.

Before concept selection, leaders should ask:

Which option has the lowest lifecycle emissions, and which has the lowest uncertainty?
Are we comparing options using consistent boundaries?
Which assumptions could reverse the ranking between options?
Are operational emissions, construction emissions, and embodied carbon all visible?

Before FEED approval, they should ask:

What are the top uncertainty drivers?
Which data are generic, supplier-specific, measured, or estimated?
What would reduce the contingency before final investment decision?
Are temporary power, logistics, marine spreads, and commissioning included?
Are future design changes being treated as contingency or planned scope?

Before sanction, they should ask:

Is the carbon budget approved, and does it include contingency?
Who can authorize contingency drawdown?
What are the escalation triggers if forecast emissions exceed the budget?
How will carbon impacts be handled in change control?
Does the financial model consider material carbon price or regulatory exposure where relevant?

During execution, they should ask:

Is contingency reducing as actual data replace assumptions?
Are supplier declarations arriving as contracted?
Are construction fuel, vessel days, temporary power, and rework being measured?
Are changes increasing emissions without equivalent approval?
Has the baseline been updated, and was that update justified?

At handover, they should ask:

What variance occurred between estimated and actual construction emissions?
What assumptions should improve the next project estimate?
Are operational emissions within the forecast range?
Are monitoring systems in place for the emissions sources that matter?

These questions are practical. They also signal that emissions performance sits inside mainstream project governance, not outside it.

A sensible way to start

Organizations do not need a perfect methodology before using carbon contingency. They do need a clear one.

A practical starting point is to define the boundary first. Decide whether the project is estimating construction emissions, embodied carbon, operational emissions, full lifecycle emissions, or a combination. Align that boundary with the GHG Protocol categories and relevant ISO approaches where appropriate.

Next, build a base estimate with a clear data hierarchy. Separate measured data, supplier-specific data, engineering quantities, generic emission factors, and judgment-based allowances. The estimate should show where confidence is high, medium, or low.

Then identify uncertainty drivers through the normal risk process. Do not ask sustainability teams to do this alone. Engineering, project controls, contracts, construction, commissioning, operations, logistics, and finance all hold part of the picture.

After that, quantify the contingency in a way that fits the project. Some teams may use ranges by category. Others may use risk-weighted allowances. Mature organizations may apply probabilistic methods, as they do for cost and schedule risk analysis. The level of sophistication should match the decision at hand.

Finally, set governance rules:

When can the contingency be used?
Who approves drawdown?
What evidence is required?
When is the baseline reset?
How is variance reported?
When should contingency reduce?
What happens if forecast emissions exceed the approved budget?

This is where many systems fail. They create a number but not a control mechanism.

The best model is close to cost-contingency governance. Some allowance sits with the project manager. Some sits with the sponsor. Some risks require board or steering committee approval. The exact structure will vary, but ownership must be clear.

For EPC contracts, carefully consider how to allocate carbon risk. Owners may control the design basis and operating philosophy. EPC contractors may control construction methods, logistics, temporary power, and supplier selection within constraints. Suppliers control product data and manufacturing choices. If contracts allocate cost and schedule risk but ignore emissions risk, disputes and blind spots are likely.

For infrastructure and utilities programs, carbon contingency can sit at both project and portfolio level. A rail, water, power, or port authority may accept that some projects will draw more allowance while others release it. That requires consistent boundaries and comparable reporting.

For mining and energy projects, the lifecycle dimension is critical. A construction-only contingency may be useful, but operational emissions from power supply, fuel use, methane, process energy, or downstream integration may dominate the real climate and commercial exposure.

So, should projects carry an emissions allowance?

Yes: where emissions matter to the project decision, major projects should carry a quantified allowance for carbon uncertainty. But call it what it is, govern it like project control, and reduce it as evidence improves.

A carbon contingency will not cut emissions by itself. It will not replace engineering judgment, supplier engagement, design discipline, or credible transition planning. It will not fix poor data if no one improves the data. Don't confuse it with offsets, targets, budgets, or planned growth.

Its value lies elsewhere. It makes uncertainty visible. It helps leaders compare options without false precision. It connects emissions to stage gates, procurement, change control, risk management, and lifecycle forecasting. It records why the forecast changed. It gives project teams a way to manage carbon uncertainty with the same seriousness they already apply to cost and schedule.

The professional takeaway is straightforward: carbon contingency is not a license to emit more. It is a tool for managing emissions uncertainty before uncertainty becomes locked-in cost, risk, and carbon. For capital-intensive projects, that makes it worth serious consideration.