I agree with JF Pragnell (NCE 22/29 July) that a debate about methods of assessing delays would be a good idea. A lot of time and money is spent in arbitrations arguing the strengths and weaknesses of the different methods now used. If the matter could be debated in principle away from the pressures of a particular case, we might come up with an authoritative recommendation which could make these arguments unnecessary.
In such a debate, I would still advocate starting the analysis with a logic linked 'as built' programme. JF Pragnell says it is very difficult to create this from records. I have never found it easy, but it is always possible. The logic derives from the nature of the work done and can be established quite firmly and, if their are no records of when the work was actually done, there are no grounds for the claimed delay anyway. Pragnell admits the alternative method works best when 'the actual progress reasonably reflects the planned progress'. I have never met this condition in a complex delay analysis.
Martin Barnes (F), Cornbrash House, Kirtlington, Oxfordshire OX5 3HF