Water companies and consultants had a very busy 2017 preparing their draft Water Resources Management Plans for 2019 (dWRMP19), and these plans will be published later this month. For Artesia it has been a very busy 12 months. Over that time we have delivered:

  • Ten household consumption forecasts;
  • Four non-household demand forecasts;
  • One set of population, property and occupancy forecasts;
  • Three option identification and evaluation projects;
  • One headroom project; and
  • One water resources demand analysis project.

These projects involve in some of the most technically challenging aspects of water resources planning – particularly when trying to forecast things that we have limited knowledge or information about (e.g. unmeasured household water use), and when forecasting uncertainty itself (i.e. headroom). We have spent significant time and effort applying the most appropriate technical approaches to these challenges but in some cases I am left wondering: was it worth the effort? Or to put this more positively: what can we learn from this experience to do it better next time!

Here are three ideas for doing things differently.

Include the most cost-effective options for managing demand

We carried out option identification and evaluation for three water companies covering four supply regions from the Midlands to East Anglia and the South East. The feasible option lists for each region were very similar and cost-benefit analysis indicated that similar options always tend to perform best. The industry should recognise that there are a handful of proven options for reducing consumption including increasing metering, household water efficiency retrofits and for leakage, pressure management and active leakage control. These options provide proven least cost savings, so they why aren’t they included in all deficit zones and maybe even in all WRMPs?

Do consumption forecasts more frequently

Wouldn’t it be great if we and/or our water company clients could update their forecasts of demand from households and non-households with a few mouse clicks! We think consumption forecasting tools should be built to enable this.

Updates to key input data would be made yearly, based on annual returns whilst the underlying trends and model parameters could be reviewed and updated in the years between WRMPs. This would reduce the pressure and workload required in the run up to the five-yearly WRMP and help ensure consistency in the forecasts over time.

 Reconsider how headroom is analysed

Most companies currently use a sophisticated method that combines several different kinds of uncertainty (often derived from detailed models) in a statistical analysis to provide a distribution of risk over the planning period. Companies then decide on a suitable level of risk they wish to include in target headroom, which generally starts high and decreases over time. Results tend to be a target headroom of between five and ten percent of deployable output.

This method is time-consuming and can only be done once all other aspects of supply-demand analysis complete. It is also increasingly difficult to reconcile headroom with new risk-based methods and understand its meaning in terms of resilience. Headroom feels like it is need of simplification and redesign to align it better with the new resilience paradigm.