Current tunneling schedules and costs are hard to estimate due to the uncertainty of geologic conditions. This leaves everyone – engineer and client alike – wondering what the final cost and end date will be. New prediction tools are being developed that better understand the geology and create a more accurate model of the time and cost.
What Methods are Currently Being Used?
The Decision Aids for Tunneling (DAT) developed in 1999 has been used to make decisions around cost and schedule. This tool is better than previous ones. It allows for the variation and uncertainty from the geology and construction methods involved. The model helps engineers understand the risk profile and flags aspects that might cause issues and delay.
The DAT tool distinguishes itself from other risk assessment tools by allowing engineers to simulate the uncertainties in geology and construction methods involved in tunnel construction. The model results provide insights on the risk profile and help to identify problematic aspects of the project.
All of the models developed so far, including the DAT one, only provide estimates prior to construction – they provide no updates as construction progresses. This can make the predictions extremely inaccurate and unhelpful when surprises pop up during construction – as they often do.
An Ongoing Predictive Tool
Universities in Iran published a solution in Tunnelling and Underground Space Technology in 2021. Their proposed tool estimates the tunnel geology and time and costs of construction – much like previous tools.
Where this tool differs is it’s use of the continuous-space, discrete-state Markov process to predict the tunnel geology. And it’s estimation of the time and costs using a Monte-Carlo simulation.
The continuous-space, discrete-state Markov process operates with the probability of each step depending on the state after the previous step. The Monte Carlo method uses random samples that are repeatedly taken. Those samples are formed into a numerical result.
The combination of these continuous updating techniques allows this tool to update as tunnel construction progresses.
According to the authors of the study, ‘It can be used when needed to update project prediction values and use the result to update schedule, resource allocation plan and financial plans of a tunnel project.’
This makes for a highly flexible tool that is extremely useful during construction.
The study authors decided to test out the tool on the construction of the Hamru tunnel in Iran. Testing of this tool involved creating an initial time and cost estimate that was updated twice – once after 200 meters had been constructed and a second time at 400 meters.
The study found that total uncertainty of time and cost was reduced by 45% and 52% after the first update and 66% and 61% after the second update. The initial prediction was off by 42 days and $0.31 million when compared to the actual values. By the second update this had been reduced to 4 days and $0.06 million.
The authors consider this a significant reduction in uncertainty. They recommend using this tool not just for creating predictions, but also to understand the likelihood of the construction time and costs deviating significantly during construction.