Impacts Of Rainfall And Water Demand Changes On Optimal Tank Size Of Rainwater Harvesting Systems

With climate change,urbanization and industrialization,water scarcity and waterlogging have become urgent problems to be solved in urban development.Rainwater harvesting systems（RHS）can collect and store rainwater resources during rainfall and replace tap water in many water demand scenarios,which can mitigate water shortage and relieve pressures on urban water supply and drainage.However,the water saving and stormwater capture efficiency of RHS have changed with the climate change and water demand change,and the uncertainty of the economic feasibility of RHS caused by this change has become an important factor limiting its widespread application.In this study,three economic indicators,i.e.,net present value（NPV）,benefit-cost ratio（BCR）and payback period（PP）were used to evaluate the economic feasibility of RHS at four cities located in different climatic zones of China with four water demand scenarios.A downscaling technique based on the CLIGEN model was evaluated and employed to generate the future daily rainfall with projections of Global Circulation Models（GCMs）.The future daily water demand was generated with exponential smoothing method and multiple linear regression model.Scenario analysis was used to set up three scenarios of water demand change with different degree of constraint.And the optimal tank size of RHS was determined with a multi-objective optimization method based on the three indicators.Then the impacts of climate change and water demand change on economic feasibility and optimal tank size were investigated by comparing results calculated using the future（2021-2050）and historical（1989-2018）daily data.The main conclusions are as follows:（1）The spatial downscaling of nonlinear regression analysis and the temporal downscaling of CLIGEN model can accurately simulate the rainfall data of the four cities from 1960 to 2005.The R² of the nonlinear regression equation of monthly rainfall ranged from 0.904 to 0.994,with an average value of 0.976.The K-S test results of measured and simulated rainfall showed that Fuzhou,Beijing,Urumqi and Yinchuan respectively had 10,8,9 and 6 months passed the test with a significance level of 0.05.The months in Urumqi and Yinchuan that did not pass the test were mainly during the non-rainy season period from November to April.（2）The multiple linear regression model,with water price,household size,income and city water saving amount as independent variables,can accurately simulate the urban water demand from 1999 to2018,and the correlation coefficient of the model is 0.898.The p value of F test is 7.254×10^-6.The results of evaluation indexes also showed that the exponential smoothing method can be used to simulate the changes of future influencing factors.（3）The downscaling rainfall data shows that mean annual rainfall in Fuzhou in 2021-2050 is going to decrease by 1%with that in 1989-2018,and in Beijing,Urumqi and Yinchuan are going to increase by66%,28%and 38%,respectively.The relative changes of indoor water demand with low,medium and high constraint scenarios are going to increase 29%,2%and decrease 25%,respectively.The outdoor water demand with low,medium and high constraint scenarios are going to increase 7.95%、8.39%and8.82%,respectively.（4）The costs and benefits of RHS in lifetime increase with reservoir volume.Three economic indicators show a two-stage changing trend that increases first and then decreases with reservoir volume increases,and the corresponding optimal volume of each index is different.Furthermore,better economic benefits can be achieved in the case of more water demand and rainfall.（5）Using the historical rainfall and water demand data from 1989 to 2018,the optimal sizes of reservoir for RHS in Fuzhou are determined to be 7 m³,10 m³,33 m³ and 14 m³ with the water demand scenarios of laundry,toilet flushing,irrigation and combined use,respectively.For RHS in Beijing,the optimal sizes are 5 m³,9 m³,15 m³ and 10 m³ with the four water demand scenarios.The optimal sizes are 5 m³ for both toilet flushing and combined use in Urumqi.No volume can make RHS in Yinchuan economically viable for any water demand scenario.（6）Changes in future rainfall and water use from 2021 to 2050 will have a positive impact on the economic feasibility of RHS in the four cities.Under the premise of economic feasibility,the maximum NPV of RHS at the four cities in the four water demand scenarios increased by 12748.77 RMB on average,the maximum BCR increased by 0.50 on average,and the minimum PP decreased by 8 years on average.This indicates that better economic benefits can be obtained in the future by designing reasonable tank size.The results also showed that the scope of volume corresponding to economic feasibility of RHS will be broaden in the future.For RHS in Urumqi and Yinchuan,some scenarios when RHS cannot achieve economic feasible in historical time will achieve economic feasible due to the changes of rainfall and water demand in future.The changes of the optimal tank size also indicated that it is necessary to build a reservoir with a larger volume in the future to achieve better economic benefits.