Short-term domestic water demand: estimation, forecasting and management

摘要: In the UK, water resource problems during droughts of 1988-1992 and well- publicized 1995-1996, serve to highlight finite nature potable resource. Demand management is increasingly considered a fundamental tool in promoting sustainable strategy. However, equal importance development accurate demand forecasts that work parallel with measures. These should predict all components ot use on different planning horizons. Presently, plcs have very limited understanding factors influencing short-term domestic demands rely upon crude methods forecasting. no precise definition exists within industry. Previous research defined ‘short- term be between twenty-four hours seven days ahead. It was argued effective weather prediction operationally useful forecast times were determining factor definition. The consumption data used this derived from two methods, (i) zonal metering (ii) individual household metering. Welsh Water Yorkshire provided data, which refers flow zone has many households. Essex Suffolk Water, Thames measure single determination of: underlying influence for both short- medium- long-term demands. The influential aided exploration modelling strategies Approaches explored included pragmatic approach, based form accounting using series 'lookup tables’, advanced approaches, including stepwise regression, without k-means cluster analysis, univariate multivariate ARMA time-series modelling. most successful approach then determine how future scenarios such as changes population base, climate, culture technology might characteristics demands. Household size property type appears exert greatest short-term, influenced by antecedent prevailing day’s conditions, day week, calendar effects, school holidays No best overall emerged one promising techniques. change demand, suggests increases are associated climate culture. widespread implementation measures potential suppress indeed reduce less than those present day.