PM 585 Week 3 Estimating Duration and Cost Example

Nine Thousand Dollars a Line, Give or Take: Estimating Duration and Cost for 2,400 Lead Service Line Replacements

[Student Name]

University of Phoenix

PM/585: Project Scheduling and Value Management

Week 3 Assignment

[Instructor Name]

[Date]

The water authority, its unit costs, crew rates and totals are composites written for a model paper.

Weeks 1 and 2 defined the scope of Allegheny Ridge Water Authority's first lead service line project and broke it into a work breakdown structure with 60 construction blocks of about 40 addresses each. The composite western Pennsylvania utility now needs estimates of how long the work will take and what it will cost, for the board, the state revolving fund and the city. This paper sets out the methods and results.

Starting From the Authority's Own Data

Last fall the authority ran a pilot of 120 replacements in two neighborhoods with one contractor crew. Pilot records show an average direct cost of $8,600 per line, ranging from about $5,900 for short lines under grass to $14,800 for long lines under concrete streets with trench restoration, and an average of 1.1 crew-days per replacement. The pilot is the best evidence available because it reflects local soils, labor rates, permit practices and housing stock. The current standard notes that estimates improve when they are based on relevant historical data and refined as more information becomes available (Project Management Institute [PMI], 2021).

A Parametric Cost Model

Pilot data showed three variables driving cost: line length, surface over the line and location under a street or under a sidewalk and lawn only. Using the inventory, each of the 2,400 lines was assigned to one of four types:

Type A, short line under lawn and sidewalk (about 38 percent of lines): about $6,800 each.

Type B, medium line crossing a sidewalk and part of the street (about 34 percent): about $9,200.

Type C, long line under a full street lane (about 20 percent): about $12,100.

Type D, line requiring interior work in a finished basement (about 8 percent): about $13,500.

The weighted average is about $9,210 per line. Multiplying by 2,400 lines gives a direct replacement cost of about $22.1 million.

Productivity and the Learning Curve

Crews in the pilot were slower at first. Wright (1936), studying aircraft manufacturing, observed that labor per unit fell by a roughly constant percentage each time cumulative output doubled, a pattern now called the learning curve. Pilot data fit a curve of about 90 percent: each doubling of cumulative replacements per crew cut the time per line by about 10 percent. Starting from about 1.4 crew-days for a new crew's first line, a crew drops to about 1.02 crew-days by its eighth line and about 0.92 by its sixteenth, then levels off near 0.9, where site conditions rather than practice limit speed.

Across all 2,400 lines the average is about 0.93 crew-days, or about 2,230 crew-days in total. At roughly 140 working days per construction season after weather losses, one crew provides about 140 crew-days a season, so the work needs about eight crews: eight crews provide 1,120 crew-days a season and 2,240 over two seasons, which matches the need with no margin at all. The schedule therefore adds a ninth crew for the second half of the first season, giving about 2,310 crew-days of capacity and a small cushion for weather and access delays.

A Three-Point Estimate for a Typical Block

For a 40-line block worked by one crew, the team estimated an optimistic duration of 34 crew-days if all lines were Type A or B and access agreements were complete, a most likely duration of 40 and a pessimistic duration of 56, allowing for several Type D lines, rain and late access. The weighted estimate is (34 + 4 × 40 + 56) / 6, or about 41.7 crew-days, with a standard deviation of about (56 minus 34) / 6, or 3.7 crew-days. Blocks are scheduled at 42 crew-days, about eight working weeks per crew including restoration hand-offs.

Rolling Up Through the WBS

Costs rolled up by WBS element:

1.1 Project management: about $1.15 million over two years.

1.2 Design and permits: about $0.88 million.

1.3 Procurement, of which materials are mostly inside unit costs: about $0.31 million for filters and administration.

1.4 Community engagement and access: about $0.62 million.

1.5 Service line replacements: about $22.1 million.

1.6 Restoration beyond trench patching: about $2.4 million.

1.7 Records and compliance: about $0.34 million.

The base estimate is about $27.8 million.

A base estimate is the cost of the project going as planned, which is the one outcome nobody should plan on.

Contingency and Reserve

Contingency covers identified risks and sits inside the cost baseline. The risk register's main cost risks are a higher share of Type D lines than the inventory suggests, asphalt price increases and lower access rates requiring more outreach visits. Their expected costs total about $1.9 million, so contingency is set at $2.0 million, giving a cost baseline of about $29.8 million. A management reserve of $1.5 million, about 5 percent, covers unidentified risks and is controlled by the board, for a total authorized budget of about $31.3 million.

Testing for Optimism

Flyvbjerg et al. (2002) analyzed a large sample of transportation infrastructure projects and reported that budgets came in too low on roughly nine projects in every ten, with average overruns that had not improved over decades, and argued that the pattern reflected optimism and, in some cases, deliberate misrepresentation. Their remedy is to compare a project's estimate with the actual outcomes of similar completed projects. The authority compared its per-line estimate with published figures from four other utilities' completed lead line programs, adjusted for region and year. Their actual costs ranged from about $8,400 to $13,200 per line for full replacements, and the median sat near $10,500. The authority's $9,210 direct cost, about $12,400 per line including all project costs and contingency, sits within that range on an all-in basis. The comparison did lead to one change: the share of Type D lines was raised from 6 to 8 percent, since other utilities reported more interior work than their inventories had predicted.

Basis of Estimates

A short basis-of-estimates document records the pilot data, line type assumptions, learning curve parameters, crew counts, season length, risk expected values and the comparison with other utilities, so that the board, the funding authority and future estimators can see how the numbers were built and update them.

Refining the Estimates

Estimates will be refined after the first 400 replacements, when actual costs by line type and crew productivity can replace pilot figures, and again before the second season, when the planning packages from Week 2 are broken into blocks.

Conclusion

The authority's estimate combines its own pilot data in a parametric model, a learning curve for crew productivity, three-point ranges for blocks, a WBS roll-up, contingency tied to risks and a reserve controlled by the board, for a total authorized budget of about $31.3 million over two seasons with nine crews at peak. A comparison with other utilities' actual results tested the estimate for optimism and changed one assumption. The estimate is a starting point that will be refined as the authority's own crews generate better data.

References

Flyvbjerg, B., Holm, M. S., & Buhl, S. (2002). Underestimating costs in public works projects: Error or lie? Journal of the American Planning Association, 68(3), 279-295. https://doi.org/10.1080/01944360208976273

Project Management Institute. (2021). A guide to the project management body of knowledge (PMBOK guide) (7th ed.). Project Management Institute.

Wright, T. P. (1936). Factors affecting the cost of airplanes. Journal of the Aeronautical Sciences, 3(4), 122-128. https://doi.org/10.2514/8.155