Here are our latest thoughts covering the legal dispute between a disappointed supplier – Energy Solutions – and the UK’s Nuclear De-Commissioning Authority (NDA). Energy Solutions were the incumbent supplier but lost a tender back in 2014 and challenged the decision, principally on the grounds of an unfair evaluation process.
Justice Fraser found in favour for the claimant, and damages may run into many millions – some estimates suggest we may be talking about £100 million here. As we said before, there are so many interesting aspects to this story, it will keep the procurement lawyers going for years and is likely to be used as case law in many challenges for years to come. Today, we will look at one aspect of the “unfairness” claim, and one that has some valuable lessons for procurement practitioners around Europe – and probably beyond.
If we look at the process followed by the NDA, the evaluation team got together and carried out a consensus scoring process, which is what we would expect for a major procurement where there are likely to be multiple evaluators. However, as we reported here in a previous article, it seemed that there was concern when a favoured supplier potentially failed to meet some threshold scores on certain questions, which could have seen them disqualified.
So the team leader went back into the AWARD eSourcing system, that was used amongst other things to collect, record and manage the scoring process, and changed the scores that had been awarded by the team. He then apparently talked to the team and made sure they were happy with the new scores.
All a little odd, most experienced procurement people would say. But what was even worse, and was something the judge really picked up on, was the fact that there was no real audit trail available to explain exactly why the scores had been changed. The judge was left with the view that this was part of the process he called “fudging” designed to make sure the bidder was allowed to remain in the process.
As the judgement says, “The system described by Mr Rankin had the result (whether by accident or design) that the most important step in reaching the final score, namely a change from a fully consensual score, to a different one, after completion of the consensus process, was a step taken with the reasoning for it recorded nowhere at all. This is contrary to the whole ethos of the evaluation process as designed”.
So what are the lessons for procurement here? Firstly, we should aim to make the consensus scoring process a single event wherever possible, with a clear set of scores as the output. A good audit trail recording how and why those scores were arrived at is also important of course.
However, there is always the chance that having been through this process, there is a desire to alter those scores. For instance, there is usually a programme board, steering group or other body that sits in the ultimate governance and decision making role for a major procurement project. My advice to that body has usually been that when the results of the evaluation are presented to them, they have the right to interrogate the evaluators and they may identify failings in the evaluation process.
If they are unhappy, they can then ask the evaluation team to go back and look again at specific aspects of the scoring. The steering group cannot or should not take it on themselves to re-score however – that would very likely lead to trouble. But the evaluation team may re-consider. The judge agreed that changes are permissible.
“To have the independent conclusion of the whole team of SMEs (the subject matter expert evaluators) on a particular Evaluation Node changed would, provided it were done in compliance with the obligations upon the NDA, be entirely permissible”.
However, if any revisions to the initial scoring are made, then the audit trail for that is absolutely critical. Instead, the “NDA’s approach was to limit the permanent record of what occurred to the absolute minimum of information.”
Indeed, the justification for such a change in our view needs to be more detailed, robust and clear than the initial scoring, because we have to explain why we marked something as maybe a 2 out of 5 first time around, and now it is going to be a 4 out of 5 (or indeed a 1).
You should assume when you record the reasons that you are explaining this to the disappointed supplier, or indeed to the judge in court. It may well be genuine – “we initially mis-interpreted the marking system that was defined in the tender”. Or “we failed to spot that the bidder had included the key charts in the appendix”. We’ve all been there in complex procurement processes.
But what you cannot do, as the NDA did, is change the scores, fail to record the reason why, hope for the best and then prove pretty much incapable of explaining that when it comes to the crunch and your day in court. The NDA’s error there really was so fundamental, it seems hard to believe that such a large organisation, with legal advisers involved, failed to do thing properly.
As we say, there are many interesting aspects to this case, and we will be back with more thoughts on other points raised shortly.