A significant amount of work was completed outside the original terms of reference. The project initially began in 1995/96, but due to protracted delays in construction, the Operations Support Component went dormant until July 2002. The project was restarted 6 months before the expected data of commissioning of the PHLC link canal and the downstream control system. A considerable number of further delays were experienced, and IWMI effectively became the "honest broker" between the various contracted parties - eventually using experience, data, diagnostic analysis and use of some of the support tools to successively fix a series of commissioning problems relating to retained structures, gate settings and so on. IWMI also took responsibility for farmer training on CBIO, but not for WUA formation, a parallel initiative which also ran way behind schedule.

A lot of work and effort was put into this project, with limited returns in the research domain. The project's efforts were much appreciated by ADB, which granted several project extensions. The final project workshop was held in Peshawar in December 2005.

Outputs of the project included:

• Web based-management information system created, with complete sample data for one larger distributory canal. NWFP staff trained on use of MIS and system handed over to the department. This MIS surprisingly generated a lot of interest in the department, beyond PHLC.
• Decision support tools created:
  1. Dynamic Programming model to optimize water allocations from Tarbela and minimise losses in electricity generation.
  2. Crop Based Irrigation Operation software completed (2003) and 30 users in NWFP irrigation department trained.
  3. CBIO trialled over 4 seasons in Maira Branch canal, with varying levels of success.
     1. CBIO training video for farmers produced in 2004: shown to 4,498 farmers in all villages within the Maira command area.
  4. Hydrodynamic models of Machai (upstream control), PHLC (link canal) and Maira (downstream control) created and used for scenario assessment of canal operation and canal capacity, using the SIC Model (Cemagref).
     1. (2 DoI and 2 IMWI staff trained in SIC in Colombo by Cemagref and IWMI, in 2003.
     2. Calibration efforts
     3. MSc Thesis - Thushari.
  5. Remote sensing analysis and GIS based determination of actual crop patterns, demonstrated and documented - incorporated into MIS.
  6. PhD analysis of canal operations at secondary level - night time closure alternatives to CBIO rotational operations.
• Analysis an monitoring of warabandi practice, groundwater use and groundwater behaviour - showing that 1) with upstream control, and partial implementation of CBIO, excess water supply was the norm resulting in 2) a breakdown of traditional rotational water distribution at tertiary level, especially at night, which 3) generated considerable return flows through the tails of both secondary and tertiary systems, and coupled with a decline in groundwater abstraction in previously low supply areas, resulted in 4) rising groundwater tables and water logging at the tails of the distributary canals from end of 2004 onwards, necessitating rice production in some instances.

This project typifies a certain type of work that IWMI has undertaken in the past, especially in Pakistan, where it is seen as a sort of SUPER (and reliable) consultant. This sort of work can bring in substantial finance, but has very limited research productivity, unless this is carefully and thoughtfully built in at design stage. After that, it is a matter of meeting the contract requirements, and funders often express the opinion that the research elements should be funded by IWMI itself, as it is not of prime interest to them. IWMI now avoids projects like this especially where the ultimate client (in this case DoI) has very little interest beyond the concrete and the contracts.