Earned Value Management (EVM) for Construction Schedules, Made Simple
What Is Earned Value Management (EVM) in Construction?
Earned Value Management (EVM) is a project control technique that integrates scope, schedule, and cost to measure performance. In construction scheduling, EVM answers three critical questions: Where are we supposed to be? Where are we actually? How much have we spent? The answers come from three basic metrics:
- Planned Value (PV) – The budgeted cost for work scheduled to be completed by a given date. Also called Budgeted Cost of Work Scheduled (BCWS).
- Earned Value (EV) – The budgeted cost for work actually performed. Also called Budgeted Cost of Work Performed (BCWP).
- Actual Cost (AC) – The actual cost incurred for work performed. Also called Actual Cost of Work Performed (ACWP).
From these, you derive performance indices and forecasts. The beauty of EVM is that it gives you early warning signs – not just whether you’re over budget, but whether you’re behind schedule and over budget, and by how much.
Key EVM Metrics: CPI, SPI, EAC Explained
Cost Performance Index (CPI)
CPI = EV / AC. A CPI less than 1.0 means you’re spending more than planned for the work done. For example, CPI = 0.85 means you’re getting only $0.85 of value for every dollar spent. A CPI above 1.0 is favourable (under budget).
Schedule Performance Index (SPI)
SPI = EV / PV. An SPI less than 1.0 means you’re behind schedule – you’ve earned less value than planned. SPI = 0.90 means you’ve only accomplished 90% of what should have been done. Note: SPI measures schedule efficiency in cost terms, not time. It’s possible to have SPI > 1.0 if you’ve completed more work than planned (ahead of schedule).
Estimate at Completion (EAC)
EAC is the forecasted total cost of the project. The most common formula is EAC = BAC / CPI, where BAC is the Budget at Completion (the total planned budget). If CPI is 0.80, EAC = BAC / 0.80 = 1.25 × BAC – meaning you’ll likely exceed the original budget by 25%. Other formulas incorporate SPI or both CPI and SPI.
Building an S-Curve from a Primavera P6 Schedule
An S-curve plots cumulative PV, EV, and AC over time. In Primavera P6, you can generate an S-curve by exporting resource or cost data. Here’s a simple method:
- Assign budgets to activities (cost accounts or resource loading).
- Run a schedule to get the baseline dates.
- Export the Resource Usage or Cost spreadsheet to Excel.
- Sum costs per period (weekly or monthly) to get PV per period, then cumulative.
- For EV, use actual progress (e.g., % complete) multiplied by budgeted cost per activity.
- Plot cumulative PV, EV, and AC on a line chart – that’s your S-curve.
- SPI > 1.0, CPI > 1.0: Ahead of schedule and under budget. Keep doing what you’re doing.
- SPI < 1.0, CPI > 1.0: Behind schedule but under budget. You’re spending less but not producing enough. Consider adding resources (if budget allows) or re-sequencing.
- SPI > 1.0, CPI < 1.0: Ahead of schedule but over budget. You’re burning cash to accelerate. Check if the pace is sustainable.
- SPI < 1.0, CPI < 1.0: Behind schedule and over budget – the classic trouble zone. Immediate corrective action needed.
- Using % complete based on duration, not physical progress. A 50% duration does not mean 50% of the work is done. Use earned rules like 0/100, 50/50, or weighted milestones.
- Ignoring the schedule baseline. EVM is only meaningful against a frozen baseline. Frequent rebaselining destroys trend analysis.
- Not updating actual costs regularly. AC must reflect real invoices and labour hours. Garbage in, garbage out.
- Misinterpreting SPI as a time indicator. SPI = 0.80 does not mean you’re 20% behind schedule in time; it means you’ve earned 80% of planned value. Actual time delay could be different due to float and critical path.
If you don’t have P6 handy, you can run these checks free in the browser with Project Assure. It parses your XER locally (nothing uploaded) and generates EVM metrics and S-curves automatically – no finance degree required.
Reading EVM Results Without a Finance Degree
EVM numbers can be intimidating, but here’s how to interpret them practically:
Also watch the To-Complete Performance Index (TCPI). TCPI = (BAC – EV) / (BAC – AC) or (BAC – EV) / (EAC – AC). It tells you the efficiency required for the remaining work to hit your target. If TCPI > 1.2, the target is very aggressive.
Common Pitfalls in Construction EVM
Bringing It All Together
EVM is not just for finance teams. Construction schedulers who understand PV, EV, AC, CPI, SPI, and EAC can spot trends early, communicate with project managers in a common language, and keep projects on track. Start with a reliable baseline, update progress honestly, and use tools that automate the math. With practice, reading an S-curve becomes second nature – and you’ll wonder how you ever managed projects without it.
Run these checks free, in your browser
Free, browser-based Primavera P6 XER schedule analyser — DCMA 14-point, GAO & NASA checks, EVM/S-curve, and forensic baseline-vs-update comparison. Nothing is uploaded; your XER is parsed locally in the browser. 3 free analyses, no card required.
Analyse your XER →Frequently asked questions
What is the difference between Planned Value (PV) and Earned Value (EV)?
Planned Value (PV) is the budgeted cost for work scheduled to be completed by a specific date. Earned Value (EV) is the budgeted cost for work actually performed. PV tells you where you should be; EV tells you where you are.
How do I calculate CPI and SPI in construction?
CPI = EV / AC (cost performance). SPI = EV / PV (schedule performance). Both are ratios: >1 is favourable, <1 is unfavourable.
Can I use EVM with a Primavera P6 schedule?
Yes. P6 supports resource loading and cost accounts. You can export usage data to build S-curves. Alternatively, tools like Project Assure parse XER files and generate EVM metrics automatically.
What does an S-curve tell me?
An S-curve plots cumulative PV, EV, and AC over time. The gap between PV and EV shows schedule variance; the gap between EV and AC shows cost variance. It gives a visual snapshot of project health.
How often should I update EVM data?
At least monthly, preferably weekly. More frequent updates catch deviations early. Always use the same baseline for consistency.