The problem with EVM on NEC4 programmes
Earned Value Management and NEC4 are not natural partners. EVM was developed in a fixed-scope, fixed-price environment — the US Department of Defense supplier base in the 1960s — where the performance measurement baseline was set at contract signature and changes were exceptional. NEC4 is built on the assumption that scope changes constantly, that compensation events flow continuously, and that the Accepted Programme is a living document updated every eight weeks.
This is not a theoretical tension. On an NEC4 programme with active compensation events, the performance measurement baseline changes every month. A standard EVM system that calculates variance against a fixed baseline will show variance that is partly genuine performance deviation and partly the effect of scope changes that have been properly assessed and agreed. Untangling the two is the central EVM challenge on NEC4 work.
The solution is not to abandon EVM — it remains the most rigorous tool available for tracking delivery performance against plan. The solution is to implement it in a way that reflects the NEC4 contract structure: a rolling baseline that incorporates agreed compensation events, performance measurement against the current Accepted Programme, and variance analysis that separates scope change from genuine performance deviation.
What NEC4 actually requires
NEC4 ECC core clauses do not mandate Earned Value Management by name. What clause 32 requires is that the Contractor submits a revised programme when: the Accepted Programme shows the wrong completion date; actual progress differs from planned; or when instructed by the Project Manager. What clause 50 requires is that the Contractor applies for interim payment based on the amount due under the contract.
EVM as a formal discipline is typically introduced through the contract data and Z clauses — bespoke additions that specify the EVMS standard the contractor must follow, the reporting frequency, the data that must be submitted, and how the performance measurement baseline is maintained. On major UK infrastructure programmes, these Z clauses increasingly reference AACE recommended practices (specifically 11R-88, the EVMS standard) or NEC's own Earned Value Management guidance.
Where EVM is not contractually mandated, it is increasingly expected. National Highways, HS2, Sellafield and major water company clients all have supply chain expectations around EVM that are enforced through the project controls and reporting requirements set at kick-off — and on the defence side, MoD CADMID Demonstration and Manufacture contracts typically write the EVMS requirement directly into the contract data, referencing AACE 11R-88 or the MoD's own EVMS guidance. A contractor who cannot produce BCWS, BCWP, ACWP and EAC in the format the client expects will fail the first controls review.
The practical implication: if you are working on a major NEC4 programme, assume EVM is required whether it is explicitly specified or not. Design your project controls system to support it from day one.
Building the performance measurement baseline on NEC4
The performance measurement baseline (PMB) on an NEC4 programme is built from the Accepted Programme — the agreed, time-phased plan for delivering the works. Every activity in the schedule needs a budget value (BCWS at completion) and a method for measuring physical progress (the earned value technique).
Choosing the right earned value technique for each activity is one of the most important decisions in setting up an NEC4 EVMS. The options range from fixed formula (0/100 or 50/50 — simple but crude) to weighted milestones (appropriate for design deliverables) to percent complete (flexible but subjective) to level of effort (for management activities that consume resource throughout the programme). Choosing percent complete for a 26-week construction activity with no intermediate milestones means progress measurement is whatever the foreman says it is. That is not earned value — it is opinion.
On NEC4 programmes with active compensation events, the PMB must be updated when a compensation event is implemented. This means the baseline is not a single fixed point: it is a version-controlled document with a clear history of when each change was incorporated and why. The version control discipline is not optional — it is what allows you to calculate performance against the right baseline and to explain variance to a client who will ask why the numbers have moved.
The four EVM metrics that matter
BCWS (Budgeted Cost of Work Scheduled) is what you planned to spend on the work planned to be done by today. This is your baseline — it comes directly from the time-phased PMB and changes only when the baseline is updated for an agreed compensation event.
BCWP (Budgeted Cost of Work Performed) is the budget value of the work actually completed. This is earned value — it is what gives EVM its name. If you planned to spend £100k on an activity and you have completed 60% of it, your BCWP is £60k regardless of what you have actually spent. This is the key insight of EVM: progress is measured in budget terms, not cost terms, so performance and cost are tracked independently.
ACWP (Actual Cost of Work Performed) is what you have actually spent on the work done to date. On NEC4 programmes this comes from the payment mechanism — the amount applied for and certified under the contract. The relationship between BCWP and ACWP is your cost performance index (CPI = BCWP / ACWP). A CPI below 1.0 means you are spending more than planned to achieve the work done.
EAC (Estimate at Completion) is your current forecast of total final cost. The simplest EAC formula — EAC = BAC / CPI — assumes future performance will mirror past performance. This is usually wrong on NEC4 programmes where compensation events and scope changes are the dominant drivers of cost movement. A better EAC incorporates the remaining work forecast at current rates plus agreed and anticipated compensation events. The difference between the formula EAC and the properly-built EAC is often the most interesting number in the monthly report.
The most common EVM failures on NEC4 programmes
The most common failure is not implementing EVM at all and producing a monthly report that shows percentage complete and spend to date as if they were the same thing. They are not. Spend to date tells you how much money has left the account. Earned value tells you what you got for it.
The second most common failure is implementing EVM but not updating the PMB for compensation events. A baseline that does not reflect agreed scope changes will show variance that looks like poor performance but is actually scope growth. This is misleading to the client, demoralising to the delivery team, and useless for forecasting.
The third failure is choosing percent complete as the earned value technique for every activity and then updating percent complete based on intuition rather than measurement. This produces EV data that is as reliable as the foreman's mood on the day of the update. If physical progress cannot be measured, use weighted milestones or fixed formula. Honest EVM with a simple technique is more useful than sophisticated EVM with unreliable inputs.
The fourth failure is producing EV reports that nobody reads because they are not connected to the decisions the programme director actually needs to make. EVM is a leading indicator — it tells you where you are heading before the variance becomes a crisis. If the EV report is not reaching the right people at the right time in a format they can act on, it is a compliance exercise, not a controls tool.