The specific objective of the project is to demonstrate the potential of analytical models to predict water-use and prioritize locations for additional water metering. This complements the Office of the Secretary of Defense directive to provide an increased understanding of water use to meet mission needs and inform infrastructure and resource investments.

This project will develop advanced analytical approaches using water data and models to better predict water use at mission-critical and water-intense facilities. Existing water infrastructure will be mapped geospatially to provide a baseline for one United States Marine Corps (USMC) installation. This project will provide Department of Defense (DoD) installation water managers cost-effective tools and methods to help target water use, risks, and to prioritize future advanced metering systems (AMS) deployment at military installations. Additionally, this project will use data from existing AMS to develop analytics to better understand water use where advanced metering is already taking place. If successful, this project will create an installation level baseline for water infrastructure, evaluate and validate models to predict water use at mission critical locations, establish a baseline of water-use installation-wide, and prioritize locations where additional AMS meters are needed to capture water use.

While the use of AMS along with analytical approaches in the commercial water sector is not new, the application of these approaches for mission-critical and water-intense DoD facilities has not been widely explored or applied. In addition, there are few commercial water applications implementing prediction analytics. The novel aspect of this project is to provide a DoD-specific systems engineering tool to capture water use, availability, and risk over various time horizons. The analytical approaches and tools developed through this study will be transferable to DoD military installations, with the initial proof-of-concept at a USMC installation.

Accurate water use predictions through models and analysis could help the DoD achieve water resilience, without the costs of implementing AMS or challenges associated with cyber security. Acquisition cost per meter (not including operational costs) can typically range between $130 to $1,895 [1, 2] which, when applied at the installation scale, can quickly increase. The USMC alone has 24 installations, each with hundreds to thousands of service connections, many of which are currently unmetered. Further, USMC installations use, on average, over 7.6 billion of gallons of water per year. This project would allow the DoD to cost-effectively predict water usage to help meet DoD and USMC water resilience and use goals, while also identifying mission-critical and water-intense facilities on military installations where cost-effective meters and AMS could be used.

References:

[1] “Federal Building Metering Guidance,” Department of Energy [Online]. Available: https://www.energy.gov/sites/prod/files/2014/11/f19/metering_guidance.pdf. [Accessed 28 February 2022].

[2] Adamowski, J., Karapataki, C. “Comparison of Multivariate Regression and Artificial Neural Networks for Peak Urban Water-Demand Forecasting: Evaluation of Different ANN Learning Algorithms,” Journal of Hydrologic Engineering, Vol. 15, No. 10, October 1, 2010.