The White House Utility District (WHUD), Tennessee’s largest geographic water utility serving 90,000 consumers and businesses, is using data to stem water loss and create savings for its customers. Detailed in a newly published paper, the district’s work began in 2015 with a dilemma: How to meet growing demand for water within the budget and capital constraints faced by municipal and mid-sized utilities?
Early projections indicated that WHUD might need to invest up to $20 million in transmission upgrades and treatment-plant expansions to meet its service commitments. Expanded capacity would also mean higher expenses in terms of energy—approximately 30% of the cost of producing water—employees, chemicals and maintenance.
Rather than launch a construction project, WHUD opted for another route—partnering with OSIsoft, Matchpoint, ESRI and Hydreka to develop a system to pinpoint underground leaks through software and smart meters. The results according to the utility:
- WHUD had an infrastructure leakage index (ILI) of 2.86 in 2012, which meant it was losing approximately 32% of its water through water main leaks.
- In less than four days, WHUD discovered a local stream was a water-main leak spilling approximately 147 million gallons a year, or enough for 2,239 homes.
- In two years, WHUD recovered $900,000 worth of water and has been able to drive its IDI down to 1.49, with a goal to further reduce water loss in the next year.
- The ‘smart meter’ approach allowed WHUD to avoid $200,000 worth of SCADA upgrades and recover $30,000 in employee time and productivity.
- The time needed to prepare reports on potential problems dropped from six hours to ten minutes.
- WHUD avoided the multi-million capital expansion. WHUD estimates that the interest payments on the bond payments alone would have come to $600,000 per year. WHUD predicts it will not need a major capital expansion until 2028.
The deluge of leaks
Recovering water from leaks costs (on average) $1.21 per 1,000 gallons in the US, according to data from Bluefield Research, or less than half the cost of traditional water ($3.90 per 1,000) and far less than newer solutions like desalination, which averages over $8.00. Leakage control is even less expensive than
encouraging consumers to conserve, according to data from the California Public Utilities Commission.
While the city sleeps
To accomplish its goals, WHUD segmented its service territory into 33 district metered areas (DMAs) with Hydreka’s Hydrins2 insertion meters and in consultation with Matchpoint. Data from this network of meters was then delivered to OSIsoft’s PI System, a software platform that collects, cleans and structures data from different devices to give engineers and technicians real-time insight into their overall operations and asset health.
Rather than try to understand consumption patterns in daytime use, WHUD monitored water consumption rates between 1 and 4 a.m., when few consumers would be awake and legitimate consumption would be at the lowest level within a DMA. If a DMA exceeded a threshold value of 0.5 gallons per minute per household during this time period, PI System data would be employed to narrow down the location of a potential leak. The data would then be placed on an ESRI ArcGIS map so maintenance crews could prioritize repairs.
“Without us knowing there was a problem in that area, we would have never been able to stumble upon that leak,” said Carl Alexander, GIS director at WHUD. “Since implementation, this has held true with potential leaks being found daily, some so small they could have gone undetected for years. We have also been able to proactively notify customers that they may have a leak, so it has really been win-win.”
“White House Utility District has demonstrated what can be accomplished through digital technology. Just as important, WHUD has shown that analytics and big data aren’t just for large utilities with extensive engineering departments. These tools are accessible to organizations of any size,” said Gary Wong, industry principal for water at OSIsoft. “Water utilities could become one of the most important test beds for analytics.”