Applying Stochastic Linear Scheduling Method to Pipeline Construction

Fitria H. RachmatBechtel Corporation, Texas, U.S.

Lingguang Song & Sang-Hoon (Shawn) Lee University of Houston, Texas, U.S.

Agenda• Linear Construction• Linear Scheduling Method (LSM)• Research Problem & Objectives• Stochastic LSM (SLSM)• Case Study

– Pipeline Construction– Data Collection– Automated Input Modeling– SLSM Modeling– Outputs

• Conclusions

Linear Construction Projects

• Characteristics– Involve a large number of repetitive activities– Activities occur in succession– Subject to uncertainty and interruptions– E.g. high-rise, pipeline, and highway projects

• Project Success– Effective project scheduling/control– Ensure continuous work flow w/o interruptions

Pipeline Construction “Assembly Line”

Linear Scheduling Method (LSM)• LSM– Designed for linear construction– 2D time-space graph– Production line = repetitive task– Line slope = productivity

• Benefits– Easily model repetitive tasks– Both time & space data– Visualize time/space buffers– Visualize work continuity

Location

Calendar

Floor 2 - 1

July 2

Formwork Rebar

Space Buffer

Time Buffer

July 1

Electrical

Interruption

Floor 2 - 2

A Demo of LSM

Section 1B

Section 2B

Traditional Bar Chart Schedule

Pour Section Layout

Schedule Delay - Elimination

1B

Floor 2

2B

Pour section layout

LSM Chart

Formwork

Concreting

Rebar

Electrical

Research Problem & ObjectivesCurrent Look-ahead Scheduling Practice

Historical dataPersonal experience

Deterministic schedule (CPM or LSM)

Proposed Look-ahead Scheduling Method

Collect actual project data

Collect actual project data

Stochastic LSM simulation

Stochastic LSM simulation

• Use real project data

• Include uncertainty

• Accurate schedules

Automated input modeling

Automated input modeling

Stochastic Linear Scheduling Method (SLSM)

• Actual productivity data collection• Automated input modeling

– Determine distributions of activity productivity

• Simulation Modeling– Simulation: a mathematic-logic model of a real

world system– A linear project can be modeled using “Project” and

“Activity” elements in SLSM

• Simulation experiments & outputs

A Case Study

• Case Study– Construction of ~130 miles of 30” pipeline

• Procedure– Data collection – Automated input modeling– Simulation models– Output schedules

Data Collection

Date

TaskStation

FootageProductivity

(ft/d)From To

9/15Stringing

5484+00

5636+00 15,000 15,000

9/16Stringing

5636+00

5705+83 6,983 6,983

9/17Stringing

5705+83

5806+00 10,017 10,017

9/18Stringing

5806+00

5972+00 16,600 16,600

9/19Stringing

5972+00

6140+00 16,800 16,800

Sample Actual Productivity Data

Automated Input Modeling• Input modeling

– Determine the underlying statistical distribution’s of an activity’s productivity rate

Automated using BestFit ®

Automated Input Modeling

Parameters for Fitted Distribution

Actual Productivity Data

Fitted distribution

Input Modeling Outputs

Task Name

Statistical Distributions

Surveying Exponential with mean =16629

Clearing Exponential with mean = 9527

Grading Normal with mean = 2874 and standard deviation = 1363

Trenching Triangular with low limit = 670, most likely = 1809, and high limit = 10720

Stringing Normal with mean = 4837 and standard deviation = 3011

Bending Beta with a = 2.3, b = 3.4, low = 670, and high = 13812

Welding Beta with a = 1.2, b = 1, low = 700, and high = 9800

Lower-in Normal with mean = 5882 and standard deviation = 3033

Tie-in Exponential with mean = 2007

Backfill Beta with a = 1.2, b = 2.9, low = 804, and high = 15758

Clean up Normal with mean = 3688 and standard deviation = 1221

SLSM Modeling• Establish a “Project” element

• Determine work scope• Add “Task” elements

• Productivity rate• Time & space buffer• Start time

Experiment & OutputsComparison of baseline schedule & simulated look-ahead

schedule

Experiment & Outputs

Uncertainty analysis of project total duration

Individual activity performance range

Conclusions

• Actual project data can be used to enhance look-ahead scheduling accuracy

• Automated input modeling makes simulation more accessible to industry practitioners

• SLSM successfully incorporates uncertainty in traditional LSM method.

19

Thank You & Questions