City of Bellevue stopped guessing its sewer cleaning schedule

The City of Bellevue maintains around 600 miles of sewer mainline. Day to day, keeping it clear comes down to a small crew.

For years the cleaning program ran on fixed intervals, and none of those intervals were backed by data. A line was cleaned every so many months because that was the schedule it had always been on. Some assets were jetted as often as once a month.

That program is not small. Around 255,000 feet of pipe goes through recurring cleaning every year, and at $0.98 a foot that comes close to $250,000 annually and delays service elsewhere. When buildup caused a problem between cleans, the team usually heard about it from a resident before they saw it themselves.

With one crew member monitoring across the whole network, every hour spent jetting a line that was already clear was an hour taken from the work that was actually waiting.

Bellevue installed the Captis Sewer Level Monitoring Solution on the assets with the worst maintenance history. Each kit pairs a Captis S2 edge device with a Kallipr Radar Sensor. The radar reads the level without touching the flow and the logger sends it back over cellular. Installation needs no confined space entry, so the units went onto live problem assets without disrupting the work going on around them.

The approach is simple. Clean the line, then watch how long the level data takes to show buildup returning. That interval is the real cleaning frequency, measured at the asset rather than carried over from a calendar.

The data gives the team:

• Continuous level readings on the highest-maintenance lines
• A measured view of how fast each line builds up after cleaning
• Early sight of anomalies, like flow that climbs when it should be falling
• Evidence to defend a change in cleaning frequency

The data coming back has reshaped how Bellevue cleans. On the assets reassessed so far, cleaning has moved from six-monthly to annual, and lines that were being jetted monthly are next for the same review. Those lines have held the longer interval without trouble, which is exactly the evidence the program never had before.

By Bellevue’s own read, they had been cleaning more often than the network needed, and that over-maintenance was costing them twice. It tied up the one crew member they have, and it was wearing the pipes out early. Repeated high-velocity jetting scours at pipe walls, so every unnecessary clean was shortening the life of an asset that was already sound. The sensors gave the team a reason to stop, and a record to back the decision.

Halving the frequency on a reassessed line takes that footage from two cleaning passes a year to one. Every pass the data removes comes straight off the annual cleaning cost. And the crew now covers more of the network at the interval each line actually needs.

The sensors also caught something nobody had gone looking for. At two locations, flow climbs in the small hours of the morning, when a sewer should be near its quietest. That pattern does not show up on a cleaning calendar, and it would not have arrived as a public complaint. It is on the record now, and Bellevue can work out what is driving it.

Establishing the true cleaning need at a single site takes several months of continuous level data. Once a site is settled, the sensor moves on to the next problem asset.

“What’s surprised me is how often the data tells us to leave a line alone. Once you’ve got months of readings behind that call, choosing not to send a crew out is a lot easier to stand behind than the schedule we inherited.”
– Richard Peckler, Wastewater Senior Engineer Technician, City of Bellevue

Outcomes

• Real-time visibility into sewer network conditions

• Safer maintenance with zero confined space entry

• Lower total cost of ownership through long battery life

• Reduced downtime and fewer site visits

• Increased network reliability and faster response to surcharge events

Plenty of collection-system cleaning in the US still runs on fixed frequencies that were set years ago and never tested against what the line actually needs. The time it takes is easy to underestimate. Crews spend hours servicing lines that were already clear, the jobs that actually needed doing wait their turn, and the cleaning itself can shorten the life of the pipe it was meant to protect.

Measuring the real interval, asset by asset, turns a cleaning calendar into a schedule the data can stand behind. For a utility running hundreds of miles of network with one crew member, that is what lets a small operation spend its time where the network actually needs it.