Project Management in Petroleum Engineering - Phase Gate Process, Earned Value Management, WBS, and EPCI Interface Management

Large oil and gas projects combine wells, subsea systems, facilities and export lines built by many contractors, and they have a long record of cost and schedule overruns; studies such as EY's 2014 review of 365 oil and gas megaprojects found that most were over budget or late. This guide covers the tools engineers use to control them: decision gates, the work breakdown structure (WBS), earned value management (EVM), interface management and critical path scheduling, with synthetic examples. It also corrects two calculation errors that appear often in EVM and schedule examples.

Phase gate project control. Gate reviews, Scope and WBS, Earned value, Interfaces.
Phase gate project control. Key concepts: gate reviews, scope and wbs, earned value, interfaces. Original PetroSmart Academy illustration; conceptual and not to scale.

1. Decision gates and front-end loading

Projects move through appraisal and concept identification, concept selection, definition (FEED) and execution, with a formal gate before each major commitment and a post-project review after start-up. At each gate the cost estimate becomes more precise (see the AACE estimate classes in our field development planning guide) and management decides to proceed, recycle or stop.

The quality of definition reached before the final investment decision is called front-end loading. Independent Project Analysis (IPA) has published research associating better front-end loading with lower cost growth and schedule slip; its FEL index is a proprietary benchmarking method. In practice, the gate review checks that business objectives, site data (soil, metocean, environmental baseline), and engineering (P&IDs, equipment lists, plot plan, safety studies) are mature enough for the estimate class claimed, and that open gaps have owners and dates.

2. Work breakdown structure

A WBS breaks the scope into deliverables, so that every cost, activity and risk belongs to exactly one element. Synthetic FPSO development, total installed cost $2,800 million: topsides 980, hull and marine 420, subsea 650, drilling and completions 520, export pipeline 85, project management and engineering 95, commissioning 50 (sum 2,800). Topsides at level 3: process equipment 285, piping 195, electrical 125, instrumentation and control 145, structures 130, mechanical completion and hook-up 100 (sum 980).

Rules: the WBS covers 100% of the scope; elements do not overlap; it describes deliverables rather than departments; control accounts sit where WBS elements meet the responsible organization, and each has a budget, a schedule and an agreed method of measuring progress (for example weighted milestones of 20% for engineering, 40% for procurement and 40% for delivery).

3. Earned value management

EVM compares three quantities at the status date: planned value (PV, budget of work scheduled), earned value (EV, budget of work actually done) and actual cost (AC).

CV = EV − AC; SV = EV − PV; CPI = EV / AC; SPI = EV / PV

Synthetic package, month 18: process equipment budget $285 million, 65% planned and 58% achieved, $175.8 million spent. PV = 185.25, EV = 165.30, CV = −10.5, SV = −19.95, CPI = 0.940 and SPI = 0.892. For the whole topsides (BAC $980 million): PV 580, EV 548, AC 562, so CPI = 0.975 and SPI = 0.945.

Estimates at completion:

  • If current cost efficiency continues: EAC = BAC / CPI = 980 / 0.975 = $1,005 million.
  • If remaining work is affected by both cost and schedule efficiency: EAC = AC + (BAC − EV) / (CPI × SPI) = 562 + 432 / 0.921 = $1,031 million. Dividing the whole BAC by CPI × SPI (giving $1,064 million) is a common error: it applies the factors to work already paid for.
  • Bottom-up: EAC = AC + re-estimated remaining cost, for example 562 + 425 = $987 million; compare it with the index-based forecasts to test for optimism.

To-complete performance index: TCPI = (BAC − EV) / (BAC − AC) = 432 / 418 = 1.033. Finishing on budget would require 3.3% better cost efficiency on all remaining work than the 0.975 achieved so far. Studies of US defence contracts found that cumulative CPI rarely improves much once a contract is about 20% complete, which is why early corrective action matters.

4. Interface management

An interface is a point where one package's output is another's input: physical (a riser connecting to a hull), schedule (a tree installed only after the well is completed), data (vendor data needed for the control system's cause and effect charts) or contractual (who supplies the spool between two scopes). Each interface is registered with an owner on each side, the information required and a need date, and agreed in an interface control document; changes go through change control.

Synthetic register at month 18: 3,847 interfaces, of which 2,916 agreed (75.8%), 485 under review (12.6%), 218 overdue (5.7%) and 228 not yet due (5.9%). Of the overdue items, 23 affect the critical path. Those 23 get escalation and weekly follow-up; the others are managed within the packages, using the float shown by the schedule.

5. Critical path

Synthetic network (durations in days): A drill and case well 140; B build subsea tree 180; C install tree 21 (after A and B); D install flowline 45 (after A); E install riser 28 (after D); F commission subsea 35 (after C and E); G build hull 420; H topsides installation and integration 90 (after G); I FPSO commissioning 60 (after H); first oil after F and I.

Activity Early start / finish Late start / finish Total float
A 0 / 140 322 / 462 322
B 0 / 180 334 / 514 334
C 180 / 201 514 / 535 334
D, E, F 140 / 185, 185 / 213, 213 / 248 462 / 507, 507 / 535, 535 / 570 322
G, H, I 0 / 420, 420 / 510, 510 / 570 same 0

First oil is at day 570 and the critical path is G → H → I. The longest subsea chain is A → D → E → F (248 days), giving 322 days of float; adding unrelated activities together (such as B + C + D + E + F) does not give a path length. If hull construction slips 30 days, first oil moves to day 600, unless work can overlap: starting integration when the hull is 95% complete (21 days earlier) would give 549 days, provided the remaining hull work does not interfere with topsides installation.

6. What a delay really costs

A delay mostly defers production rather than losing it, so its cost is not the daily revenue multiplied by the days of delay. For a synthetic project with an NPV of $1,500 million at 10%, shifting all cash flows by one week reduces NPV by about 1,500 × (1 − 1.10−7/365) ≈ $2.7 million, and by about $11.7 million for 30 days. To this must be added the extra costs of the delay (vessel and rig standby, project team, financing) and any production truly lost, for example if a licence or host facility has a fixed end date. Multiplying 35,000 bbl/d by price and days would overstate the cost of a one-week slip several times over.

References

  • PMI, A Guide to the Project Management Body of Knowledge and Standard for Earned Value Management
  • AACE International Recommended Practices 18R-97 and 10S-90 (cost engineering terminology)
  • EY, Spotlight on oil and gas megaprojects (2014)
  • E.W. Merrow, Industrial Megaprojects (Wiley, 2011)
  • D.S. Christensen and S.R. Heise, "Cost performance index stability", National Contract Management Journal (1993)

Related reading and community

Questions or field experience to share? Join the PetroSmart Telegram group, watch worked examples on the PetroSmart Academy YouTube channel and follow us on Facebook.

Scroll to Top