We investigate the historical urban sociometabolic water demand for developing cities with the intention of illuminating how the environmental impact of water supply has changed in cities as technology has progressed. The goal of this work is to understand the conditions under which urban development is ‘sustainable’ relative to the costs of providing water. We do this by investigating the environmental and societal costs of supplying water as demand has grown, for four cities: Singapore, Singapore; New York City and Los Angeles, United States of America; and London, United Kingdom. We establish the elasticities between model parameters and characterize past uncertainties using historical observations. We then use our System Dynamics model to perform sensitivity analysis and Monte Carlo simulation to compare the sustainability of future water resource management scenarios under a variety of plausible but uncertain future development pathways.


Singapore’s rapid population and economic growth since gaining independence from Great Britain in 1963 has put pressure on its limited water resources. Despite receiving an annual average of over 2m of rainfall (nearly twice the global average), Singapore, a small island nation off the coast of Malaysia, is among the world’s most water stressed nations due to its large population (5.4 million people in 2013), limited area (island area = 700 square km), and lack of freshwater lakes and aquifers [1]. Historically, most of the water consumed on the island has been imported from Malaysia; Singapore’s stated goal is to close this gap between water consumption and infrastructure production capacity by 2060 using a combination of demand reduction policies with expansion of desalination and reclamation capacity [2]. 

Historical data from Singapore’s Public Utilities Board (plotted in Figure 1) give evidence that recent investments in water infrastructure on the island have begun to close the gap between demand and production capacity [3, 4]. However, by how much Singapore will have to increase its current capacity to meet future demand in 1960 is an estimate fraught with uncertainties: uncertainties in estimates of population size, residential water demand, commercial and industrial water demand, and effectiveness of conservation and efficiency policies [5]. Moreover, since desalination and water reclamation are both rapidly changing technologies, the relative sustainability of these water sources in terms of cost and energy intensity, is highly uncertain [5]. When compared with the uncertainties inherent in more conventional water resources, which depend upon precipitation collected from a watershed, which mix of strategies is ‘better’?

Research Question

In this project, we are investigating the question: 

What mix of source expansion & demand reduction technologies and policies reduces the risk of urban water scarcity within the context of future uncertainty?


It has been argued that technology improvements (especially material efficiencies and extraction capacities) will offset the increasing environmental impact associated with growing population and affluence [6]. Whether they have or haven’t, whether they do or don’t, whether they will or won’t: these are still fundamental and open questions in sustainability science [7]. Past investigations have focused on characterizing the relationship between resource efficiency and socioeconomic growth but often do not consider the environmental and societal cost of providing the resource or the potential for society to mitigate through innovation, thus providing limited potential to inform international policy and design.

This knowledge gap has arisen in great part because most of the resources consumed in cities have been sourced internationally, which complicates drawing direct comparisons between environmental limits and consumption. This situation is fortunately not true for water. While much has been made about the virtual global water trade, the fact remains that many urban activities still rely on the real thing, which they consume in large volumes. These large volumes require physically intensive infrastructure, which limits the economic viability of transporting water large distances such that water is still largely a regional resource.


We characterizing the relationship between water demand, supply, and the cost of supply over time for each case study using historical parameters pertaining to sociodemographic trends, climate, infrastructure, and water policy. On the supply-side, we take into account parameters affecting the regional water balance and the financial and energy costs of collecting, conveying, treating, and distributing that water. We also allow for urban water supply to invest in reservoirs, desalination plants, and reclamation plants to expand supply in the present and future. The associated costs we assume to depend on variability in climate, infrastructure deterioration, energy intensity, learning rates, and economies of scale.

For each parameter we use time series and econometric analysis, such as moving averages, to establish normative trends in historical demand-parameters related to sociodemographic changes and socioeconomic activity as well as for costs of infrastructure. We also establish probability distribution functions for the historical variability in normative trends within parameters. We use these results to calibrate our previously developed System Dynamics model, adapted to investigate competing water production technologies and uncertainty. The characterized uncertainties provide the base case ranges for parameters that are then explored with Monte Carlo simulation. Interactions between the economy and population are captured in this System Dynamics model with the Cobb-Douglas dynamic model.

Fig. 1 Singapore’s Water Production Capacity

Example: Singapore 

We used 50 years of historical data on Singapore’s water use, infrastructure production capacities, and financial expenditures to calibrate our System Dynamics model which were gleaned from Singapore’s Public Utilities Board (PUB) Annual Reports, Master Plans and Census data [3]. Figure 1 depicts data from PUB Annual Reports on Singapore’s production capacity for freshwater.

Fig. 2

The System Dynamics model was calibrated using information about changes in household size, household affluence, and population as well as information about the water intensity of the economy and changes in the economy to endogenously predict water demand [4]. The calibration results (Figure 2) gave us confidence to examine the vulnerability of water availability in Singapore to future scenarios (Figure 3). 

Fig. 3

Figure 3 shows some selected results from sensitivity analysis for different scenarios involving the percent reduction in the economic intensity of water, which was defined as the amount of water in cubic meters used to generate a dollar of GDP (units of cubic meters per dollar GDP). For reference, the average over 50 years is about 4% reduction each year. Water availability was shown to be most vulnerable to when there is more than one system stress: for instance, when the water intensity of the economy decreases to slowly and there are changing precipitation patterns due to climate change (blue); or water imports are reduced before sufficient capacity has been built (black). Figure 4 shows the major watersheds on Singapore.

Fig. 4 Singapore's major watersheds


[1] Gassert, Francis and Reig, Paul and Luo, Tianyi and Maddocks, Andrew. (2013). Aqueduct Country and River Basin Rankings: A Weighted Aggregation of Spatially Distinct Hydrological Indicators. (Working Paper). Washington, D.C.: World Resources Institute. < http://www.wri.org/sites/default/files/aqueduct_coutnry_rankings_010914.pdf >

[2] Tortajada, Cecilia. (2006). Water Management in Singapore. Water Resources Development, 22 (2): 227-240.

[3] Public Utilities Board. (1965-2009). Annual Report. Singapore: Ministry of the Environment & Water Resources.

[4] Noiva Welling, Karen. (2011). Modeling the Water Consumption of Singapore Using System Dynamics. (Master’s Thesis). Massachusetts Institute of Technology, Cambridge, MA.

[5] Wong, Melanie. (2013). Flexible Design: An Innovative Approach for Planning Water Infrastructure Systems Under Uncertainty.  (Master’s Thesis). Massachusetts Institute of Technology, Cambridge, MA.

[6] Steinberger, Julia K. and Krausmann, Fridolin. (2011). Material and Energy Productivity. Environmental Science & Technology. 45: 1169-1176.

[7] Sekulova, Filka and Kallis, Giorgos and Rodríguez-Labajos, Beatriz and Schneider, Francois. (2013). Degrowth: from theory to practice. Journal of Cleaner Production. 38: 1-6.

[8] Khan, S. and Khan, M.A. and Hanjira, M.A. and Mu, J. (2009). Pathways to reduce the environmental footprints of water and energy inputs in food production. Food Policy, 34: 141-149.