No description available.

Anthropogenic climate change is arguably the greatest challenge of modern times posing significant risks to natural resources and the environment. Socio-economic change, severe droughts, and environmental concerns focus attention upon sustainability of water supplies and the ability of water utilities to meet competing demands worldwide. The 2012 Climate Change Risk Assessment identified water security as one of the most significant climate threats facing the UK. It is now recognised that household water demand management could offer a low regret adaptation measure (both financially and environmentally) given large uncertainties about future climate and non-climatic pressures. This thesis uses Anglian Water Services (AWS) Golden 100 dataset to explore the climate sensitivity of historic micro-component water-use. This work contributes to a larger integrated assessment of the South-East England water system under the EPSRC Adaptation and Resilience to a Changing Climate Coordination Network (ARCC CN). The Golden 100 is a metered record of 100 households daily water consumption by basin, bath, dishwasher, external, kitchen sink, shower, WC and washing machine use. The archive also includes socio-economic information for each household, dates of the year and daily time series of observed minimum temperature, maximum temperature, sunshine hours, soil moisture deficit, concurrent, and antecedent rainfall amounts. The methodology developed within this research provides a portable approach to error trapping, formatting and mining large, complex water sector datasets, for exploring the relative sensitivities of micro-component metered water-use to weather/non-weather variables. This research recognises both the importance of the choice to use a micro-component and the volume used. As such, logistic and linear generalised regression techniques are employed to explore the relative sensitivity of these two aspects of water-use to climatic and non-climatic variables. The 2009 UK Climate Projections (UKCP09) projections and climate analogues are then used to bound a climate sensitivity analysis of the most weather-sensitive micro-components using temperature and rainfall scenarios for the 2050s and 2080s. This research provides empirical evidence that the most weather sensitive micro-components are external and shower water-use. A key contribution of this research to existing knowledge is the non-linear response of likelihood and volume of external water-use to average air temperatures. There is an abrupt increase in the likelihood of external water-use on days above ~15??C. Climate sensitivity analysis further suggests that by the 2080s, under a hotter/drier climate, average unmetered households could be 8% more likely to use external-water and expend ~9 litres more per day during the summer. For the same parameters, high water users (defined here as the 90th percentile) could consume ~13 litres more external water per day. Importantly, this research has re-affirmed the relative importance of behavioural drivers of water-use as manifested by pronounced day of week and bank holiday signatures in both the likelihood and volume of use statistics. As such, this prompts future studies and water management efforts to consider the impact of behavioural drivers as well as climate. It must be recognised that the small sample size of the Golden 100 combined with the Hawthorn effect, self-selection and sample biases in factors such as socio-economic status, billing method and occupancy rate all limit the sample representativeness of the wider population. As such, any predictions based on the data must be treated as illustrative rather than definitive. Furthermore, the results are probably specific to the demographic and socio-economic groups comprising the sample. Nonetheless, this research sheds new light into water-use within the home thereby adding value to a dataset that was not originally collected with household-level, weather-related research in mind.

Thesis (MEng)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: The focus of the study is to theoretically assess tank-water demand and employ methods to establish the actual tank-water demand at selected houses in a case study area. This study also examines the influence of domestic rainwater harvesting systems when used in combination with a municipal water distribution system. The case study comprises of 410 low cost housing units in the Western Cape. The system demand patterns of low cost housing units are uncharacteristic, when compared with suburban system demand patterns, and cannot be defined by traditional models. Similarly, the use of rainwater harvesting systems in these areas follows an unconventional routine that is yet to be defined. A stochastic end-use model for water demand is developed which produces temporal profiles for water supplied from both sources, namely the water distribution system and the rainwater harvesting system. The model approximates a daily system and tank-water demand pattern for a single domestic household, using @RISK software. The demand estimation methodology is clarified through application on a particular case study site where harvested rainwater is frequently utilized. Estimates of the parameter values are based on consumer surveys and previous studies on the case study area, where the household size was defined in the form of a probability distribution. The results confirm the atypical system demand patterns in low cost housing units units. Although two clear peaks exist in the morning and in the evening, a relatively constant average flow is present throughout the day. A sensitivity analysis of all the model parameters verified that the household size has the most substantial influence on the tank-water demand pattern. The system and tank-water demand patterns were compared to published average daily water demand guidelines, which confirmed that increased water savings could be achieved when the rainwater source is accessible inside the household with minimal effort. The stochastic demand profiles derived as part of this research agree with the metered system demand in the same area. The results of this study could be incorporated into the future development of national standards. / AFRIKAANSE OPSOMMING: Die fokus van die studie is om die tenkwater-aanvraag teoreties te ontleed en metodes in werking te stel om die werklike tenkwater-aanvraag vas te stel by geselekteerde huise in ‘n gevallestudie area. Hierdie studie ondersoek ook die invloed van plaaslike reënwater-herwinningstelsels wanneer dit gebruik word in kombinasie met ‘n munisipale waterverspreidingstelsel. Die gevallestudie bestaan uit 410 laekoste behuisingseenhede in die Wes-Kaap. Die stelsel-aanvraagpatrone van laekoste behuisingseenhede is verskillend wanneer dit met voorstedelike stelsel-aanvraagpatrone vergelyk word en kan nie gedefinieer word deur tradisionele modelle nie. Soortgelyk volg die gebruik van reënwater-herwinningstelsels in hierdie areas ‘n onkonvensionele roetine. ‘n Stogastiese eindgebruikmodel vir water-aanvraag is ontwikkel, wat tydelike profiele genereer vir water wat van beide bronne verskaf word, naamlik die waterverspreidingstelsel en die reënwater-herwinningstelsel. Die model bepaal by benadering ‘n daaglikse stelsel- en tenkwater-aanvraagpatroon vir ‘n enkele plaaslike huishouding, deur @RISK sagteware. Die aanvraag-beramingstegnieke word verduidelik deur toepassing op ‘n spesifieke gevallestudie, waar herwinde reënwater gereeld gebruik word. Die parameter waardeberamings is gebaseer op verbruikers-opnames en vorige studies oor die gevallestudie-gebied, waar die grootte van die huishoudings bepaal was in die vorm van 'n waarskynlikheidsverspreiding. Die resultate bevestig die atipiese stesel aanvraagpatrone in laekoste behuisingseenhede eenhede. Alhoewel twee duidelike pieke in die oggend en die aand voorkom, is ‘n relatiewe konstante vloei dwarsdeur die dag teenwoordig. ‘n Sensitiwiteitsanalise van al die modelparameters bevestig dat die grootte van die huishouding die grootste beduidende invloed op tenkwater- aanvraagpatrone het. Die stelsel- en tenkwater-aanvraagpatrone was vergelyk met gepubliseerde gemiddelde daaglikse water-aanvraag riglyne wat bevestig dat meer waterbesparings bereik kan word waar die reënwaterbron binne die huishouding beskikbaar is met minimale moeite. Die stogastiese aanvraagprofiele, wat as deel van hierdie navorsing afgelei was, stem saam met die gemeterde stelsel-aanvraagpatroon van dieselfde area. Die resultate van hierdie studie kan in die toekomstige ontwikkeling van nasionale standaarde opgeneem word.