Contractor claims:- using Monte Carlo to assess them

Copy of article posted by Kelly O’Brien on Linkedin.

Recently, I was asked by a client to use a risk based approach to support it in assessing and settling a large number of contractor claims. The subject project was in the construction sector and there were over 100 claims in total from the contractor. It was the first time I did a task like this. It went much better than I expected and is the reason why I wish to share my experience.


We isolated one very large claim relating to COVID-19 and dealt with that separately. All the others we put into a register (excel). We gathered the client side project team comprising design, project management, contract and cost specialists and we went through each claim item, one by one, in a dedicated workshop.  For each claim item, we assessed the probability of the contractor being successful if it went to a formal dispute and also what range of costs the contractor could be entitled to. This was done on a 3 point basis; low, most likely and high. We also added in a commentary setting out the basis of each assessment the team made. Some claims were more straightforward than others to assess, the ones with a time dimension being the most challenging.

Probability of success

The probability of success aspect of the assessments was interesting. The team considered if each of the contractor’s claims had merit. The construction contract, a design and build, was referred to frequently. Where the contract was clear on the employer being at risk on the subject matter of a claim item, then a probability of 100% was applied. In some cases, however, even where an obligation or risk was 100% with the contractor under the contract, and clearly so from my interpretation albeit I’m not an expert, we assumed a probability of 10 or 15%. This was based on the advice and experience of our contract specialist, his opinion being that one can never pre-empt how an adjudicator would assess a claim.  

Monte Carlo

After the workshop was finished, and some outstanding cost assessments undertaken, we ran a Monte Carlo on all the claim items. We used the standard P50P90 model for this which is freely available. This produced four figures for the client, P10, P50, P80 and P90. The P10 to P90 figures were in effect the reasonable boundaries within which the client could settle while demonstrating that value was being retained or achieved. The eventual settlement was within this range which I feel is a great achievement. The negotiation back and forth with the contractor took about five weeks from the date of the workshop. I had no involvement with this. The contractor never got sight of the team’s assessment.


For me, the process worked for a number of reasons. First and foremost, a settlement was reached in a non-adversarial way. Working relationships with the contractor were maintained which is important as construction of the project remains ongoing. The key staff members of the client side team were involved throughout and jointly owned the exercise. The assessment was client driven and was undertaken in a structured way. It was very thorough. There was a methodology, audit trail and report out supporting the assessment. This also helped with the internal approvals to enter into an agreement with the contractor. Was this a win-win outcome? The parties involved believe that it was.