3 Prediction of demand
For a major water supply project there is a substantial time between the recognition of the need for water and the completion of the project (the lead time), so predictions of the future demand for water are essential and need to be made for around 25 years ahead. Prediction starts by looking at how the demand for water has varied in the past. Figure 2 shows how abstractions varied between 1971 and 2001 in England and Wales. Although there is little difference in total abstractions for 1971 and 2001, there were large variations during this period, with a maximum of around 45 × 106 m3 per day in 1992, and a minimum of around 32 × 106 m3 per day in 1994.
Which use in Figure 2 has had the least variation between 1971 and 2001?
The public water supply, which varied only between around 14.5 × 106 m3 per day in 1971 to around 17.5 × 106 m3 per day in 1990.
The other uses have much greater variability; agriculture, for example, has increased from very little before 1983 to around 5 × 106 m3 per day from 1992.
The variation in which use has had the greatest effect on total abstractions?
The electricity supply industry; for example, between 1992 and 1994, the amount of water it used dropped by about 9 × 106 m3 per day.
Comparable data by use do not exist for other parts of the UK, but there are data for the public water supply (Table 2).
What was the general trend in the amount of water used by the public water supply in the UK between 1990 and 2002?
It has generally decreased, from over 20 × 106 m3 per day in 1990/1 to around 18.5 × 106 m3 per day in 2001/2. However, the decrease is not regular; there are yearly increases, particularly in 1994/5 and 1995/6.
Table 2 Public water supply in the UK: 1990/1-2001/2, in 106 m3 per day. (DEFRA, 2004)
|England and Wales||17.38||17.20||16.76||16.76||17.11||17.32||15.98||15.34||15.33||15.26||15.78|
Box 1 Water metering
[Figure 3 With a water meter, a customer is charged for the volume of water used.] With a water meter (Figure 3), a household, office or industry pays for the water it actually uses instead of paying a fixed charge unrelated to use. Most homes in Europe and North America have water meters but in England and Wales, although almost all new homes are installed with water meters, in total only 24% of homes are metered (2003). This is generally higher in areas of water stress, for example, the Anglian water company region meters about half of its domestic customers. Offices and industry are usually metered.
The advantage of metering is that it tends to reduce demand as it raises awareness and encourages consumers to consider and value their use of water. In the long term, it has led to changes in attitude to water, so that more people are choosing devices that are water efficient. This encourages manufacturers to make and advertise water-efficient household appliances, such as WC cisterns, washing machines and dishwashers.
The disadvantages are the cost of metering and a concern about health. Installing meters in existing homes costs about £130-£200 per home, an enormous investment across the country. There is also a concern on health grounds that some households will sacrifice cleanliness in order to economise.
The EA long-term strategy for water resources in England and Wales (EA, 2001) is that 50 to 75% of households should be metered by 2025.
Public water supplies in England and Wales generally increased between 1971 and 1990 (Figure 2). Domestic consumption rose during this period due to population growth and the increasing domestic use of water per person. The fall since 1990 (Table 2) is due to domestic metering, more efficient use of industrial water, periods of industrial recession, and reduced leakage. The metered water supply in England and Wales rose gradually through the 1990s, and the unmetered supply fell, partly due to the increase in domestic metering.
Graphs such as Figure 2 have to be projected far into the future because of the long lead times necessary in planning for new water resources. As well as looking at past trends, prediction of the future demand for water involves breaking down the total present demand into domestic, industrial and agricultural components, and identifying the economic, social and population factors which are likely to affect each of them in the future.
Forecasts include assumptions about increases in domestic demand due to greater use of appliances such as automatic washing machines and waste-disposal units, and decreases in domestic demand due to more showers and fewer baths, dual-flush WCs, and water metering to houses. They also include assumptions about population growth, the level of economic activity, climate change (Section 5) and the rate of leakage from the system (Box 2).
Box 2 Leakage
More water is lost through leaks in the public water supply distribution system than is put to any one use. In 2002/3 the leakage in England and Wales was estimated as 3.6 × 106 m3 per day, around 22% of the water put into the system. The leakage rate is greater in cities, where water mains date back to Victorian times, most of which are now dilapidated. Here leakage can reach 40%. However 'lost' is a relative term, as much of the water that leaks from the mains recharges aquifers.
Water is lost through continual gradual leakage, as well as from spectacular temporary bursts caused by vibration, soil compaction, corrosion or excavation when installing gas pipes and electricity and telephone cables (Figure 4). There are around 20 bursts per year in England and Wales for every 100 km of water mains — and there are over 3 × 105 km of mains.
Leakage can be, and is being, reduced by replacing or relining old mains, but it is both very expensive and very disruptive, involving digging up roads and tunnelling under buildings. Mandatory leakage control targets were introduced during the 1990s in England and Wales. This reduced a leakage of around 5.1 × 106 m3 per day in 1994/5 (about 31%) to 3.6 × 106 m3 per day (22%) in 2002/3. Leakage reduction at present (2004) is planned to be 1.5% a year. The Thames water company, which has many Victorian mains in the London area, managed to reduce its leakage from 1.1 × 106 m3 per day in 1995/6 to 0.77 × 106 m3 per day in 1998/9. However, despite continuing mains replacement, in 2002/3 this had risen to 0.93 × 106 m3 per day, a leakage of 33%.
As well as being slow, there is also a technical and economic limit for leakage reduction. The aim is to reach an economic level of leakage, which is the point at which the cost of reducing leakage is the same as the value of the water saved. This point is not fixed because of improving technology in locating leaks and changes in the price of water.
The assumptions that have to be made to predict demand reveal the uncertainties inherent in forecasting, and the forecasts may turn out to be highly inaccurate (Table 3).
Table 3 Past forecasts of public water supply for England and Wales.
|Year of forecast||For year||Predicted demand /106 m3 per day||Water supplied /106 m3 per day|
This can be seen easily in retrospect: for example, the 1971 England and Wales forecast for 1981 was 19.2 × 106 m3 per day, whereas the real demand was much less — about 15.9 × 106 m3 per day. The 1973 estimate for the year 2000 of 28 × 106 m3 per day is now seen to be far too high. For a shorter prediction period, the 1987 forecast for 1991 was accurate, but time will tell whether its predictions for 2011 or the 1992 predictions for 2021 are any good. Judging by previous forecasts, they probably will not be. The failure in recent years to make accurate predictions of the future demand for water in the UK has been partly caused by the difficulty in predicting industrial changes. 2001 predictions are discussed in Section 6.