Favicon of Rockwell Automation

EnWin Utilities

EnWin Utilities Reduces Water Main Breaks 21% Saving $125,000 Annually with MPC

21%Water Main Break Reduction
$125,000Annual Savings
2.8 psiSystem Pressure Reduction
EnWin Utilities
Metric Before After Impact
Annual Water Main Breaks 238 187 21% reduction
Annual Savings $0 $125,000 $125,000 generated
Average System Pressure −2.8 psi vs baseline 2.8 psi reduction
Pressure Standard Deviation improved 29% improvement

The Challenge

EnWin Utilities operates the Windsor Utilities Commission in Windsor, Ontario, distributing approximately 48,000 million liters of water annually to more than 72,000 customers. By late 2012, the utility was averaging 238 water main breaks per year at a cost of roughly $5,000 per incident — a chronic drain driven primarily by pressure spikes during pump start and stop events. The aging distribution network, characterized by iron water mains, corroding pipes, and soil erosion, is particularly vulnerable to such fluctuations. Because water is a non-compressible fluid, any pressure change propagates system-wide. Existing PID-based control logic could only manage a single variable at a time, leaving operators unable to coordinate pressure, flow, and drive speed simultaneously — and unable to prevent escalating infrastructure damage.

The Solution

EnWin partnered with Rockwell Automation to deploy a Model Predictive Control (MPC) solution in two phases. Phase One, commissioned in June 2013, introduced a server-based MPC controller alongside 17 remote pressure stations distributed across the service area, integrated with the existing Allen-Bradley ControlLogix SCADA platform and PowerFlex variable frequency drives. The controller replaced single-variable PID loops by managing multiple simultaneous inputs — pressure readings, drive speeds, and flow data — to hold system pressure at the lowest viable level for adequate service. Phase Two, completed in January 2014, moved MPC logic onboard the ControlLogix controller itself, eliminating the need for a dedicated server. This reduced control interval speed from 15–16 seconds to under one second, enabling real-time coordination of medium voltage drives and adjustable flow control valves to offset pressure spikes the moment a pump starts or stops.

Results

EnWin's shift from reactive repair to predictive pressure management produced measurable outcomes within the first year of operation. Annual water main breaks fell from 238 to 187 — a 21% reduction — generating approximately $125,000 in annual savings from reduced repair activity, lower electricity consumption, and decreased system leakage. Average system pressure dropped by 2.8 psi, and pressure standard deviation improved by 29%, indicating significantly more consistent delivery across the distribution network. Pump start and stop pressure spikes — the primary driver of main breaks in aging infrastructure — were virtually eliminated after Phase Two commissioning. The onboard MPC architecture also removed the capital and recurring licensing costs of a dedicated server, further reducing operational expenditure.

Key Takeaways

  • Model Predictive Control outperforms single-variable PID in aging water distribution systems by coordinating pressure stations, drives, and flow control valves simultaneously rather than chasing a single setpoint.
  • A phased rollout — server-based MPC first, then onboard — reduces deployment risk while giving operators time to build confidence before full system integration.
  • Embedding MPC directly in an existing PAC eliminates dedicated server and licensing costs, lowering total cost of ownership beyond the operational gains alone.
  • Consistent pressure management is measurably more cost-effective than reactive pipe repair programs; controlling the pump start/stop event addresses the root cause, not the symptom.
  • Utilities cannot replace aging infrastructure fast enough to keep pace with deterioration — advanced process control can bridge that gap cost-effectively.

Share:

Details

AI Technology
Predictive ML
Company Size
SME
Quality
Verified

Have a similar implementation?

Share your customer's AI results and link it to your vendor profile.

Submit a case study →