What a Walla Walla water operator monitors while 35,000 people turn on their taps
The City of Walla Walla announced that administrative offices would close for Labor Day, September 7. Water and wastewater operations would continue as usual. Somebody had to work the holiday. Somebody has worked every day since July 16, when the Paradise Fire started burning inside the watershed that supplies roughly 85 to 90 percent of the city's drinking water.
The governor declared an emergency over the fire's threat to that supply and the flooding expected to follow it. The city said it had observed no impact to water quality. Both statements were accurate. Neither one tells you what's happening in the control room.
The creek and the fire
Mill Creek comes down out of the Blue Mountains through a municipal watershed and feeds the city's treatment plant, rebuilt in 2019. By late August, the Paradise Fire had burned more than 8,000 acres inside that watershed. As of September 9–10, it stood at 9,655 acres and 28 percent containment, with crews shifting from active suppression toward repair.
The plant matters more than the acreage. Walla Walla runs an unfiltered surface water system. Most cities push their water through a full filtration sequence — coagulation, settling, sand filters — that can absorb a wide range of raw-water conditions. Walla Walla's plant uses sedimentation basins, a coarse roughing filter of sand and gravel, ultraviolet light, and chlorine. That's adequate for what Mill Creek normally delivers. What Mill Creek normally delivers is not what a burned watershed produces after a rainstorm.
The city said so itself: the treatment plant is not equipped with filtration capable of handling large-scale wildfire-related changes to the water. Which leaves the question of what the operator is actually watching for, and how close it is.
The number
Turbidity. That's the number. It measures how much suspended material — sediment, ash, burned organic matter — is floating in the water, expressed in nephelometric turbidity units, or NTU. Washington law sets the ceiling at 5.0 NTU for raw water entering an unfiltered system before disinfection. Exceed that, the source comes offline. Do it repeatedly and the city can lose its permission to operate without conventional filtration, a permission that once revoked would require a capital project nobody has budgeted.
Below the ceiling there's a second mark at 1.0 NTU. Cross it and the operator has to pull additional coliform samples within 24 hours. Coliform bacteria are the standard proxy for contamination; they don't necessarily make you sick, they indicate that something capable of it may have gotten in. On an ordinary summer day the operator is already collecting at least five raw-source fecal-coliform samples a week on separate days. Cross 1.0 often enough and the lab work stacks up.
5.0 NTU: source must come offline. 1.0 NTU: additional coliform sampling required within 24 hours. The operator's actual decision point — the reading that would prompt a call to switch to wells — isn't in any public document.
In practice that means pulling water from the creek, carrying it to the plant's state-certified microbial lab, and setting cultures in incubators that won't tell you anything for at least eighteen hours. While those sit, continuous turbidity readings scroll across the SCADA screens, the supervisory control system that reports in real time what every sensor in the plant and the distribution network is doing. Chlorine residual gets checked with a colorimeter. The roughing filter and the settling basins need eyes on them. The job description lists turbidity instruments, chlorine colorimeters, water baths, incubators, programmable logic controllers.
I stood watches at sea for five years, and the part nobody mentions is that a watch is mostly boredom with arithmetic in it. You take readings you fully expect to be normal, and you take them anyway, because the entire value of the exercise is that somebody was looking during the twenty minutes when they weren't. Each instrument in that plant measures one thing. The operator reads them against each other: turbidity edging up while the fire map adds acreage, creek flow shifting while demand holds flat, yesterday's culture finally reporting on conditions that have already changed. Weekly standby assignments mean somebody is always reachable. During a fire burning inside your own watershed, reachable and relaxed are different conditions.
So 5.0 is the wall and 1.0 is where the workload escalates. But the operator's real decision point — the reading, or the rate of change, or the combination of fire behavior and creek conditions that would make the call obvious — appears in no public document. Not in the 2020 Water System Plan. Not in the Water Shortage Response Plan the City Council adopted in April. The city has said it will keep using Mill Creek "as long as the water can be treated to meet all drinking water standards," and may switch preemptively if fire retardant threatens the creek.
The city says it has increased water-quality testing since the fire began. It has not published the turbidity readings, the analyte list, the sampling calendar, or the lab reports. The most recent Consumer Confidence Report covers calendar year 2025, which is to say before any of this. "No observed impact" is an assurance with no number under it. The margin between current readings and the trigger could be a mile or it could be a hair.
The usual complaint about public data runs the other direction, an agency publishing a figure so broad that nobody can act on it. Here the operator almost certainly has a working number, and what reaches the public is the legal ceiling and a reassurance. The data exists. It sits in the control room.
The wells, such as they are
If Mill Creek goes offline, the city falls back on seven deep wells drawing from the Columbia River Basalt aquifer, drilled between 1941 and 1971, running from about 789 to 1,412 feet. Combined installed-pump rating: 24.6 million gallons per day.
That number deserves the kind of scrutiny you'd give a used-car odometer. The Water System Plan warns, in its own pages, that pump ratings describe installed equipment and not demonstrated aquifer production. On August 27 the city said six wells were functional and could fully supplement the system. It didn't say which one was down. It didn't publish a current capacity test.
| Measure | Gallons per day |
|---|---|
| Combined pump rating (7 wells) | 24.6 million |
| Estimated 6-well capacity | 20.28–22.44 million |
| Average daily consumption (Aug 24–30) | 13.06 million |
| Historical max-day demand (2011–17) | 20.6 million |
On the hottest day of the hottest week, every sprinkler in town running, the low end of six-well capacity barely clears the historical peak. That assumes each well produces at nameplate, which assumes the aquifer cooperates. Oregon's water resources agency has documented declines of up to 150 feet in basalt aquifers across parts of the Walla Walla Basin since 1940, pumping running ahead of natural recharge. Those are regional figures, not measurements of the city's own wells. They do describe the system those wells draw from.
Each well has a history, and the histories read like a shop log on a truck fleet that should have been turned over ten years ago. Well 5's casing failed during cleaning in 2022, letting 27 feet of clay into the hole. A fire damaged Well 2's electrical panel in March 2023, and city records at the time described that well as important during high turbidity, peak summer demand, and loss of surface water. Well 7 has sulfur, taste-and-odor problems, and water that comes up around 80 degrees. The city's capital plan schedules fixed generators and automatic transfer switches for Wells 1 and 6, bids due September 24, meaning the backup power project was still out for bid when the fire started. Condition assessments, rehabilitation, treatment evaluation are all queued behind it.
The city has said its deep-aquifer supply can meet demand for ten years or longer. That's a planning assertion. No current pumping test, no drought stress test, no model of six-well sustainable yield under continuous emergency operation has been made public.
Who makes the call
You'd expect this part to be simple. The Water System Plan puts the system under the City Manager, with day-to-day authority delegated to the Public Works Director. The Water Shortage Response Plan sets up a Water Management Team to evaluate conditions, a Public Works Director who contacts the City Manager, and a City Manager who determines the response stage. But that plan's thresholds are about production against demand: gallons available versus gallons needed. It has nothing to say about water that is physically available and chemically ruined.
No public document assigns any specific position final authority over a water-quality-driven switch from Mill Creek to the wells. The city's water page names Adrian Sutor as Water Operations Manager and Joe West as Water Treatment Supervisor. Certified operators monitor quality, adjust treatment, and rotate weekly standby with call-back availability.
So the public record gives you an organizational chart and a set of reporting requirements. What it doesn't give you is who, at two in the morning on a Saturday after a hard rain, can shut the intake and start the wells: the operator on shift, the treatment supervisor by phone, or somebody further up who has to be woken. The state requires notification within 24 hours after certain events, including water above 5.0 NTU delivered to customers, disinfection failure, or emergency source shutdown. That's a report filed afterward. It says nothing about who was authorized to cause it.
The storm
The fire receding doesn't mean the water problem is ending. It may mean it hasn't started.
A peer-reviewed analysis of post-fire water quality research found that 91 percent of studies measuring suspended solids or turbidity recorded increases after wildfire, with reported magnitudes running from modest changes to more than 25,000-fold. Those changes tend to show up with the first significant runoff. A 2025 study of 538 western watersheds found sediment delivery persisting up to eight years, peaking years after the fire in places where shallow landslides arrive late.
The governor's declaration named post-fire flooding, landslides, and debris movement specifically. No soil-burn-severity map or Burned Area Emergency Response assessment for the Paradise Fire had been released through September 10. Without that assessment, nobody can model what the first real rainfall puts into Mill Creek.
The city finished a Watershed Resiliency Plan in June 2026, weeks before the fire started. It treats wildfire as a long-term watershed threat. It is a planning document, not a switch protocol.
Walla Walla has gone to wells before. During the 2019 plant commissioning the city ran entirely on well water. In February 2020 flooding took the intake offline. In July 2023 there was a planned maintenance switch. The physical act is practiced. None of those involved months of ash and sediment loading. The 2020 Water System Plan discusses the 2005 School Fire, which burned 78 square miles north of the municipal watershed, not inside it. No wildfire has yet forced this city off Mill Creek as its primary source.
The watch
The operator is watching for something that hasn't happened here, with instruments calibrated for normal operations, against a decision threshold that isn't publicly defined, backed by well capacity that's been asserted rather than demonstrated, waiting on a rainstorm that could come next week or next January.
The taps run clean, and there's no reason to doubt it. Thirty-five thousand people experience their water system as a faucet handle. The operator experiences it as a set of readings and a margin that gets narrower without anybody outside the plant being able to see it.
Labor Day came and went. The offices reopened Tuesday. The turbidity readings from the past two months remain unpublished, the burn-severity map hasn't been released, and the first hard rain on eight thousand burned acres hasn't arrived. The notice said operations would continue as usual.
- Post-fire sediment persistence: A multi-basin study found that wildfire-driven water-quality changes in western U.S. watersheds can persist for up to eight years, meaning Walla Walla's source-water problem could outlast any single fire season regardless of containment.
- Well backup power procurement: The city's bid solicitation for generators and automatic transfer switches at Wells 1 and 6 closes September 24, so the status of emergency power at the city's two aquifer-storage-and-recovery wells should become clearer by late fall.
- Regional aquifer decline: Oregon's water resources agency has documented pumping-driven groundwater declines across the Walla Walla Basin's basalt aquifers since 1940, a long-term constraint worth tracking if the city's wells shift from supplemental to primary supply.
- Smoke remediation standards in California: Two bills awaiting the governor's signature would direct California to develop statewide wildfire-smoke remediation standards for standing homes, a parallel case where the absence of an agreed threshold leaves households trapped between competing institutional definitions of "safe."