Unravelling industrial sector water use: Developing a conceptual model to quantify global thermoelectric and manufacturing water uses
MetadataShow full item record
PCR-GLOBWB is a gridded global model which simulates water resource availability and sectoral water use. One of the sectors, industrial, is comprised of the manufacturing and thermoelectric sub-sectors. The model initially used lumped input industrial water demand data, but this study altered the spatiotemporal precision of the model by separately calculating the sub-sectoral industrial contributions. Manufacturing water is mainly used to manufacture products, while thermoelectric water is used to cool thermoelectric power plants. Through the development of a conceptual model to estimate national manufacturing water withdrawals, and the integration of a global thermoelectric power plant dataset, the input parameters were adjusted. The conceptual model was created by calculating manufacturing water withdrawals using three equations for a reference year, and then forecasting and backcasting these values using annual technological changes. The national gross manufacturing water demand values were then downscaled by population density and coupled with an existing dataset of industrial return flows to estimate net manufacturing water demand. The manufacturing water demand was then further coupled with the gross and net thermoelectric water demand dataset. By aggregating the gross and net demands, an updated industrial water demand dataset was created. This dataset accounts for the gross and net water demands for the industrial sector at a 5 arc minute spatial resolution, and is theoretically more spatially precise than previous datasets because it incorporates the thermoelectric water demand at exact locations. Initially, PCR-GLOBWB used a downscaled industrial sector, but now that technique is only used for the manufacturing sub-sector. The newly created industrial water demand file was then integrated into PCR-GLOBWB, and the model was simulated between 1980 and 2014. Using the model results, several analyses were performed on the output to assess the effect of the updated industrial water demand dataset. Comparisons were made between the initial and updated industrial demand input files, the supply and demand of industrial water, and the PCR-GLOBWB output for abstractions and discharge. These analyses showed that adjusting the spatial precision of the input data did influence the calculated local and national industrial water scarcity, as well as the industrial water abstractions and discharge. The updated industrial gross demand input file did not greatly affect the total national demand from the initial input file, but did affect temporal trends due to the influence of the technological changes in the conceptual model. In the United States, an analysis of the national data showed that the updated model improved the temporal simulations due to the technological changes. At more local scales, however, the updated model was not as effective at estimating industrial water use. Upon further analysis, it was found that this was due to discrepancies between the input thermoelectric water demand, and the observed thermoelectric water withdrawals. Although the national scale values are improved, more research will need to be done to improve the spatial precision of the model locally.