"The programme" means different things at different stages. Using the wrong level of detail at the wrong time causes problems: a tender programme too thin to price from, or a contract programme so detailed that every change becomes a dispute.
The tender programme
A tender programme supports a bid or a funding decision. It needs to show a credible overall duration and the logic behind it, not every activity. It should set out:
- key dates and the overall sequence
- the critical path and the assumptions that drive it
- regulatory hold points, such as pre-commencement conditions and Building Safety Act gateways
- major lead times, such as cranes, frames and façades
- the main risks to the duration
The contract programme
The contract programme is the baseline the project is measured against. It needs to be logic-linked and detailed enough to assess delay.
Under the NEC4 Engineering and Construction Contract, the programme has particular weight. Clause 31.2 sets out what it must show, including:
- the starting date, access dates, Key Dates and Completion Date
- planned Completion
- the order and timing of the contractor's operations and of work by others
- float and time risk allowances
- health and safety requirements
The Project Manager has two weeks to accept a submitted programme or give reasons for not accepting it. If there is no accepted programme, one quarter of the price for work done can be retained until a first programme is submitted.
JCT contracts are less prescriptive, but the master programme is still the reference point for extension of time claims, so it deserves the same rigour.
The working programme
Working programmes and short-term look-aheads are for running the site. They break the contract programme into weekly and daily activities by trade and area. They change constantly, and they should, as long as they stay consistent with the contract programme.
Keeping them connected
The most common failure is three programmes that do not agree. The tender programme assumed a sequence the contract programme abandoned, and the working programme follows neither. A single logic-linked model, reported at different levels of detail, avoids that. It is also what funders and monitoring surveyors want to see.