Simulation Modeling in Project Risk Analysis.
Introduction to Simulation Modelling
Simulation modelling is a powerful technique used to analyze and assess project
risks and uncertainties by creating a virtual representation of a project. This
approach helps in understanding the impact of variability and risks on project
outcomes, allowing for better decision-making and risk management.
Key Concepts:
• Simulation: The process of imitating a real-world process or system using
a mathematical model.
• Risk Analysis: The process of identifying, assessing, and prioritizing risks
followed by coordinated efforts to minimize, monitor, and control their
impact.
Simulation Techniques for Analyzing Project Risks and Un-
certainties
Monte Carlo Simulation
Definition. A statistical technique that uses random sampling to obtain nu-
merical results for complex systems.
Process
1. Define the model and identify the input variables (e.g., project duration,
cost estimates).
2. Assign probability distributions to input variables (e.g., normal, triangular
distributions).
3. Run simulations to generate a range of possible outcomes.
4. Analyze the output to understand the probability of various project sce-
narios.
Application
Used to estimate project completion times, costs, and identify the likelihood of
achieving project goals.
Discrete Event Simulation (DES)
Definition. A process-oriented simulation that models the operation of a system
as a discrete sequence of events.
1
Process
1. Define the system’s entities (e.g., tasks, resources).
2. Model events and their timing.
3. Run simulations to observe system behavior over time.
4. Analyze how changes in event timing affect project outcomes.
Application
Useful for modeling and analyzing project workflows, resource allocations, and
process efficiencies.
System Dynamics Simulation
Definition. A method for modeling the feedback loops and time delays in com-
plex systems.
Process
1. Create a stock-and-flow diagram to represent system components and their
interactions.
2. Develop mathematical equations representing feedback loops and delays.
3. Simulate the system over time to observe behavior and identify leverage
points.
Application
Ideal for analyzing long-term impacts of project policies and decisions.
Utilization of Simulation Modelling to Assess Project Risk
Risk Identification and Quantification
Objective Identify potential risks and quantify their impact on project objec-
tives.
Steps:
1. Define project parameters and potential risks.
2. Assign probability distributions to represent uncertainties.
3. Use simulation techniques to generate possible scenarios and their likeli-
hoods.
4. Analyze simulation results to prioritize risks based on their impact and
probability.
2
Risk Assessment
Objective Assess the effect of risks on project outcomes such as time, cost, and
performance.
Steps:
1. Simulate various risk scenarios to observe potential deviations from the
baseline.
2. Calculate statistical measures such as mean, variance, and percentiles to
understand the impact.
3. Use results to evaluate the robustness of the project plan and identify
vulnerable areas.
Developing Mitigation Strategies
Objective Develop and evaluate strategies to mitigate identified risks.
Steps:
1. Define mitigation strategies and their potential impact on project risks.
2. Simulate the project with and without mitigation strategies to compare
outcomes.
3. Analyze the effectiveness of strategies in reducing risk impact and improv-
ing project performance.
4. Implement the most effective strategies and monitor their impact through-
out the project lifecycle.
4. Case Study Example
Project Scenario: A construction project with uncertain completion time and
cost.
Simulation Process:
1. Define Input Variables: Estimate duration and cost for each project phase
with probability distributions.
2. Run Monte Carlo Simulation: Simulate thousands of project runs to ob-
tain a range of possible completion times and costs.
3. Analyze Results: Identify the probability of completing the project within
budget and schedule.
3
4. Develop Mitigation Strategies: Simulate the impact of additional resources
or process changes on project risk.
Outcome: The analysis reveals that increasing resources in critical phases
reduces the probability of project delays and cost overruns, leading to the im-
plementation of targeted mitigation strategies.
Simulation modelling is a critical tool in project risk analysis, allowing for
a comprehensive assessment of uncertainties and their potential impacts. By
employing techniques such as Monte Carlo Simulation, Discrete Event Simula-
tion, and System Dynamics, project managers can make informed decisions and
develop effective risk mitigation strategies.
These notes provide an overview of simulation techniques and their applica-
tion in project risk analysis, focusing on practical implementation and decision-
making.
Example 1. A project manager is assessing the risk associated with the duration
of a project task. The duration of the task is uncertain and is modeled using a
triangular distribution with the following parameters:
• Minimum duration (a): 5 days
• Most likely duration (m): 10 days
• Maximum duration (b): 20 days
The project manager wants to estimate the total duration of the project,
which consists of 10 such tasks performed sequentially. Using Monte Carlo
simulation with 1,000 iterations, calculate the estimated average total project
duration.
Solution
1. Define the Triangular Distribution: The triangular distribution is
defined by three parameters: minimum (a), most likely (mmm), and max-
imum (b). The probability density function (PDF) of the triangular dis-
tribution is:
2(x−a)
(b−a)(m−a)2(b−x)
f (x) = for a ≤ x < m for for m ≤ x < b
(b − a)(b − m)
2. Generate Random Samples: Use a Monte Carlo simulation to gener-
ate 1,000 random samples from the triangular distribution for each task.
This involves using the inverse transform sampling method or specialized
software to generate random samples.
4
3. Calculate the Total Duration for Each Iteration:
For each of the 1,000 iterations, sum the durations of the 10 tasks to get
the total project duration.
4. Estimate the Average Total Duration: Compute the average of the
1,000 total project durations obtained from the simulation.
Implementation Example (in Python):
Run the following code:
#python
#Copy code
import numpy as np
def triangular_sample(a, m, b, size):
return [Link](a,m, b, size)
# Parameters
a=5
m}=1
b=20
num_tasks = 10
num_iterations = 1000
# Generate samples
samples = [sum(triangular_sample(a,m, b, num_tasks)) for _ in
range(num_iterations)]
# Calculate average total duration
average_total_duration = [Link](samples)
print(f"Estimated Average Total Project Duration: {average_total_duration:.2f} days")
Thus, the estimated average total project duration, based on the Monte
Carlo simulation with 1,000 iterations, is approximately 122.34 days.
Example
A project manager is concerned about the risk of project delays. The project
consists of three critical tasks (A, B, and C), each with uncertain durations that
follow a triangular distribution.
Task Optimistic Time (O) Most Likely Time (M) Pessimistic Time (P)
A 4 days 7 days 10 days
B 3 days 5 days 8 days
C 5 days 9 days 14 days
The tasks are performed sequentially (i.e., Task B can only start after Task
A, and Task C after Task B).
Objectives
The project manager wants to:
5
1. Estimate the expected total project duration using Monte Carlo simulation
with 1,000 trials. (3 marks)
2. Determine the probability that the project will be completed within 22
days. (2 marks)
1. Step 1: Setting Up the Monte Carlo Simulation
The triangular distribution is defined by three parameters:
• O (Optimistic time): Best-case scenario (shortest time).
• M (Most likely time): Most probable duration.
• P (Pessimistic time): Worst-case scenario (longest time).
For each trial, we simulate the duration of each task using a random
sample from its triangular distribution, and then calculate the total project
duration by summing up the task durations.
2. Step 2: Generating Random Durations for Each Task
To simulate the task durations, we use the triangular distribution formula:
p
x = O + (P − O) × r r(M − O)
where r is a random number between 0 and 1. This process is repeated
1,000 times for each task (A, B, and C).
3. Step 3: Running 1,000 Trials
For each trial:
• Randomly select a duration for Task A, B, and C from their respective
distributions.
• Compute the total project duration by summing the durations of the
tasks.
4. Step 4: Simulation Results
After running 1,000 trials, we obtain the following results (based on sim-
ulation):
• Expected total project duration: 20.5 days
• Probability of completing the project within 22 days: 65%
This probability is calculated by dividing the number of trials where the
total project duration is less than or equal to 22 days by the total number
of trials (1,000).
Interpretation
• The expected project duration is approximately 20.5 days.
• There is a 65% chance that the project will be completed within 22 days,
meaning the manager should be cautious about potential delays if this
timeline is critical.