The system turnaround playbook

You just took over a water system that has been falling apart for years, and you are the one who has to fix it. Violations are stacked up, nobody knows where the maps are, the board has not met in months, and every repair got pushed off until it turned into an emergency. The instinct is to fix everything at once.

That instinct is wrong. A system this far behind did not get here from one bad decision, and it will not get fixed in one good week. What looks broken (the violations, the water loss, the deferred maintenance) is rarely the real problem. The real problem is usually financial, organizational, or cultural: nobody trusts the board, staff are burned out or undertrained, rates have not moved in a decade, and every week is a new fire instead of a plan.

There is a defensible order for working through this, built from real system rescues rather than theory. It runs governance first, then survival controls, then money, then delegation, with communication running underneath all of it. Skip a step or take them out of order, and the fixes do not hold.

What does a broken system actually look like?

A broken system usually shows the same handful of signs: no certified operator, missing records or maps, outstanding violations, water loss around 30 percent, and no rate increase in ten years or more. You know it when you see it, but those signs are symptoms, not the disease.

One real case involved an 800-connection private system where the owner died without a will. The operator of record disappeared, bank accounts froze, and records for state regulators went missing entirely. Only one of five wells was still online.

Another case was a 170-connection district losing 30 percent of its water through worn-out pipes, with no reserve to replace them. Different triggers, same pattern: governance and cash had already broken down long before anyone noticed the pipes.

Where do you start?

You start with governance, before anything technical, because nothing else you do will stick until someone is clearly in charge. Define who is responsible for what and fill vacant board seats. Set term limits or require training, and hold meetings that are open, regular, and actually happen.

The first-week questions are simple: who is in charge of what, are the board seats filled, and is the public engaged at all? How this starts depends on how bad things already are.

In the 800-connection case, the state regulator stepped in and appointed new leadership because no one was left to ask. In most systems it starts smaller: a retreat or an outside-led meeting to get the board and staff on the same page about what is actually going on.

What keeps the system running while you fix the rest?

A short list of survival controls keeps the system running, not a promise to prevent every future problem. You cannot stop every emergency in week one. What you can do is make sure the system survives the next one.

A few concrete moves before the next one hits:

• Certify a backup operator, so one person disappearing does not take the whole system down.
• Write a basic standard operating procedure; even a single page beats nothing.
• Put passwords, vendor contacts, and sample schedules somewhere more than one person can find, a shared binder or file works fine.
• Start debriefing every incident: what went well, what went badly, and whether it could have been prevented.

Ask your board this directly: if the operator disappeared today, who could take a sample, and who has the keys? If nobody has a good answer, you are ready for the current emergency, not the next one.

How do you triage the money?

You protect the functions that keep water flowing, pause everything else, and start rebuilding a reserve. Broken systems are usually broke too.

Financial triage runs in a fixed order:

• Fund payroll and basic operations and maintenance first: operator and admin pay, chemicals, and power.
• Pause capital projects unless they are truly critical; a new treatment building can wait, a failed disinfection system cannot.
• Build an emergency reserve, even a small one.
• Raise rates or find short-term funding, then communicate clearly why.

A modest increase, on the order of an extra $10 per connection per month in one real case, can stop the bleeding without shocking customers. Pair it with plain communication about why, and a visible plan for what the money buys. Stabilize first, then build.

Rates alone will not always be enough. For the smallest systems, a rate increase can be necessary without being sufficient. If the customer base is shrinking, look for outside subsidy or a partnership alongside the rate fix, not instead of it.

When do you bring in outside help?

You bring in outside help as soon as governance is stable enough to manage a contract, not before. You do not have to do all of this in-house, and trying to is how one person burns out and the recovery stalls. Common things to hand off: asset or task management and billing software with a customer portal, contract operations for backup coverage, and rate studies or policy work from a technical assistance provider.

Contract out for consistency, not to lose oversight. Keep regular board check-ins and put expectations in writing.

Partnering with a nearby system is its own decision, on a spectrum from light to heavy: mutual aid for emergencies only, shared purchasing or staff, buying water wholesale, joint planning for new supply, contracting out operations for a set term, and full consolidation at the far end. Distance matters: physical mergers, sharing a source or interconnecting pipes, tend to pencil out within a few miles, while non-physical arrangements like shared services can work within about an hour's drive. Run the analysis to decide whether to partner, not to justify a decision you already made.

How do you sequence repairs, records, and rates?

Repairs wait, records rebuild as you go, and rates move early but modestly. Big capital repairs pause during triage unless they are a safety or compliance emergency; you protect core operations first and revisit the capital list once governance and cash are stable. Records rebuild in that same window: the SOP, the passwords, and the vendor contacts are the emergency version of a full asset inventory.

Once you are past triage, a complete inventory of what the system owns is the formal version of that same work. Rates move early and small, then get revisited properly once you have real numbers. A useful longer-arc tool here is a short strategic plan: pull together a team of operators, elected officials, an accountant, someone who understands the finances, and a few community members. Write down your goals and values, inventory your assets, and honestly assess your technical, managerial, and financial capacity, often shortened to TMF.

Use what you find to choose and fund the next round of options, then review the plan at least once a year. If you cannot measure whether an action worked, you cannot tell if you are actually fixing anything.

What does "fixed" actually look like?

A fixed system runs boring: no more surprises, because the basics are covered. The exit test is not zero problems, it is a routine that holds without you personally catching every ball. The board meets on schedule and can answer who is in charge of what.

Someone besides the operator has the keys and could take a sample tomorrow. Rates get reviewed on a set cadence instead of avoided for a decade. The plan you built to stabilize things gets revisited and updated instead of filed away and forgotten.

You do not have to do all of this at once, and you do not have to do it alone. You do have to start, in the order above, and keep going until the fires stop being a surprise.

For the rate and reserve side of this work, see /guides/asking-for-money-grants-rates-and-reserves. Once your board is meeting on a real cadence, /guides/the-60-second-briefing-talking-to-your-board covers how to keep those meetings useful. And when you are ready to write down exactly what the system owns, /guides/what-your-water-system-owns walks through the inventory step in more detail.