By Vladimir Moya Quiroga Gomez, Dr. Eng., Tyrone Oglesby, GISP, and Zachary Berry, GISP  

In recent years, tenants within Camp Foster, Japan, have reported low water pressure and stagnation problems. Many military family housing units on the base are near the end of service life and require major renovations. Others are due for refurbishment. The installation is presently under renovation with new infrastructure being planned and constructed.Flags of the U.S. and Okinawa, Japan

A crucial aspect of developing and maintaining a good quality of life is access to safe, reliable, affordable, and sustainable supply and distribution of drinking water. Meeting these commitments embraces multiple challenges, which requires the right combination of skills, experience and technology. Hydraulic models are a vital tool for solving such challenges. They also can verify system capacity or analyze the effect of modified infrastructure within the context of the entire water distribution system or its sub-systems.
Engineers at Camp Foster developed a geographic information system (GIS)-based water distribution model for improving the maintenance, capital improvement, and monitoring of the water distribution system. The calibrated model satisfactorily simulates the water distribution system’s hydraulics, exposing the strengths and weaknesses of the system.

Located on the southwestern coast of Okinawa, Camp Foster is located 12-km northeast of Naha. It is the largest of multiple camps on Okinawa that comprise Marine Corps Base Camp S. D. Butler, the 5th Element of III Marine Expeditionary Force. Most of its infrastructure dates to the post-World War II era, built between the 1950s and 1960s.

Engineers at Camp Foster developed a geographic information system (GIS)-based water distribution model for improving the maintenance, capital improvement, and monitoring of the water distribution system.

The water distribution system is also divided into three zones: Upper Foster, the Headquarters Area, and Lower Foster. Each is hydraulically separated by properly located valves. Two main sources provide water. A connection to the Okinawa line sends water to Upper Foster and the Headquarters Area. A connection to a water treatment plant in Camp Lester handles the Lower Foster area.Topographically, Camp Foster is split in three zones. Lower Foster is located at the west side. This is a flat zone with an average elevation of 1.5-m above sea level (masl). The second zone is a 5 percent slope where the elevation increases to 55-masl. The third (at the east) has an average elevation between 55-masl to 65-masl with some isolated hills that are 5-m to 10-m higher.


Geospatial Consulting Group International and Environmental Science Corp. carried the task to set up and calibrate a GIS-based hydraulic model for Camp Foster’s water distribution system. The work was performed using InfoWater by Innovyze, which runs as an extension inside the GIS software ArcGIS Desktop. InfoWater simulates the water distribution network by solving a matrix system that represents the mass balance at each node and the Bernoulli equation at each pipe.

Water work

The first step involved gathering pertinent documentation, historical water demand data, and GIS data regarding the location and connectivity of the main elements (pipes, tanks, pumps and valves). Information regarding pressure and current demand was collected between October 2016 and January 2017. This data was used to calibrate the model considering current conditions. Additional data from fire hydrant pressure tests helped validate the model under stressed conditions (high flow demand). This was important to better understand the model’s strengths and limitations.

Based on the statistical characteristics and assuming a 95 percent confidence, the simulated pressures are within a confidence range between -3.4 ±1.9-psi. Considering the observed pressures are higher than 40-psi, and in several cases above 80-psi, the confidence range can be assumed as a good one. The errors are due to uncertainties associated with the model: coefficients affecting head losses, reduced diameter due to pipe age, or variations in the boundary conditions such as hourly changes in water demand.

A sensibility analysis was performed to improve the understanding of the influence of uncertain sources within the model’s capabilities and limitations. Unknown water demand was the most uncertain element. The sub daily water demand variations may be responsible for pressure differences of up to 30 percent. For instance, water pressure during low demand may be more than 15-psi higher than during peak demand. The differences were confirmed by additional in situ data collection; pressure at given locations was collected at noon (average demand) and in the morning (high demand) where more than 150-psi differences were observed.

The primary limitation when calibrating was the coarse temporal resolution and scarce availability of water flow and water demand data. Although standard assumptions proved a close representation of reality, they may lead to overestimation or underestimation during peak demand. Data collection campaigns are being performed to reduce water demand uncertainties. This new data will allow for improving the model and creating more detailed simulations.


The system was evaluated considering sub daily water demand variations. The most important water demand conditions are the average daily water demand and the theoretical maximum hourly demand. Emergency conditions were evaluated assuming theoretical single-fire and multi-fire events. Maximum hourly demand and fire demand are important because they describe the system under stressed conditions.

Water pressure was evaluated considering the minimum water pressure. The GIS integration allowed creating thematic maps for an easy visualization and identification of critical areas, such as pressure nodes or head losses in pipes. During normal conditions, the system shows good water pressure. Areas with low water pressure were identified during maximum demand and during emergency conditions (fire). This behavior is typically associated with capacity problems. Small pipes experience high head losses during high flows. Such low water pressure areas may experience even lower pressure in the future due to a planned increase in demand for many of the housing areas on the installation.

Due to the growing number of requests, the distribution model team is developing a scenario management application to document workflow and to route key management responsibilities of each requested scenario.

Although the model can simulate water quality, water quality simulations require additional data usually not available. Water quality was evaluated based on age, which is a major factor in quality. As the water travels through the system it undergoes physical and chemical processes that affect quality. Based on simulated velocities and pipe lengths, the model can estimate water age. Water age analysis allowed visualizing the hidden importance of adequate tank sizing. Different storage volumes of proposed tanks will affect water age. Larger tanks may have long retention time, resulting in excessive aging. Original fresh water entering the tank may already have several days old water by the time it enters the pipe network.


Results have been encouraging. The GIS-integrated user interface allows for easy simulation and evaluation of new scenarios, new infrastructure, different operation rules, changes in water demand, and actions during emergency conditions. The model can be used as a decision support tool for evaluating different “what if ” scenarios. A number of stakeholders have started requesting simulations of several new scenarios.

Due to the growing number of requests, the distribution model team is developing a scenario management application to document workflow and to route key management responsibilities of each requested scenario. The team will focus on technical analysis, summarizing the main benefits and limitations of each requested scenario. Scenario management will be the link between the modelers who operate the model and the decision-makers who rely on it to make informed decisions.

Personnel at Camp Foster will continue working to improve water distribution services, ensuring access to a safe, reliable, affordable and sustainable supply of drinking water.

Vladimir Moya Quiroga Gomez, Dr. Eng., is Civil Engineer – Numerical Modeler, Environmental Science Corp.;

Tyrone Oglesby, GISP, is Senior GIS Analyst, and Zachary Berry, GISP, is GIS Analyst, Geospatial Consulting Group International. They can be reached at toglesby@; and

[Article originally appeared in the September-October issue of The Military Engineer.]