Sign up for our newsletter
What we found, on which network, and what it cost to find it. Roughly monthly, and never a figure we cannot show you the working for.
A distribution network is run by five people who each see a different part of it. One sees losses, one sees alarms, one sees risk, one sees twenty years, one sees the number that goes to the board. Below is how each of them describes the job. These are their words, from our research, not personas we invented in a workshop.
Ask any of the five what they would not say out loud in a meeting, and you get a version of the same answer: I don't fully trust our own data. Not the sensors, not the model, not the figure going into the board pack. That is the problem underneath all the others, and it is the one we built for.
Owns the water-loss programme, from the target to the crew that gets sent out.
Hidden leaks run for days before anyone notices. The monthly water balance is too slow to act on. A broad acoustic survey burns a week ruling out one district. And every false positive spends a specialist crew you do not have.
Runs the network hour by hour, through SCADA and telemetry.
Alarm floods end in Acknowledge-All. The data sits across SCADA, GIS, work orders and spreadsheets. You know something changed, but cause, location, severity and confidence arrive too slowly to act on.
Owns the data, the integrations and the security of critical infrastructure.
OT data was collected for control, not for analytics. Every system names things differently. New tools arrive asking for custom exports and brittle integrations, and the security review is already short-staffed.
Owns risk-based renewal and the capital plan. ISO 55000, twenty-year horizons.
The asset register is half spreadsheets and half tribal knowledge. GIS, CMMS, SCADA and leak history do not speak to each other, and what the field learns rarely finds its way back into the plan.
Accountable for the whole utility, to a board, a regulator and the public.
Ageing infrastructure is the sector's first problem and the funding gap is widening. Which makes another pilot that will not scale, or another dashboard nobody opens, an expensive way to lose a year.
It also cannot be audited, handed over, or asked a question at two in the morning when the person who holds it is asleep.
A utility with five leak teams runs them the same way every year: area by area, in a fixed order, until the network is done and the first area is stale again. Every new leak waits its turn. Move one team off that rota, and the arithmetic of the whole programme changes.
times you cover the whole network in a year. A leak that starts the day after a team leaves an area waits for the next pass.
and the other teams have walked the whole network once, clearing the 20% that was already in the ground. It ends because the team on the alarms stops new leaks taking their place.
How long a new leak runs, from the day it starts.
91× shorter, and it is the same crew doing it.
One leak, at 4 l/s. That is 346 m³ a day into the ground, every day nobody knows it is there.
of water that never leaves the network, from moving one team and hiring nobody. That is 66,000 leak-days removed and 22.8 million m³ kept — if every one of those leaks ran at the rate you set. They will not: yours will be a spread, and the small ones run longest. Put your own average in and the shape of the answer holds.
And then the part worth planning for. Once the baseline is down and one team holds the new leaks as they appear, searching harder stops paying. That is when the other teams are worth more on the work that stops leaks happening at all: condition assessment, renewal, pressure management. The programme changes shape rather than ending.
These five priorities genuinely conflict. Software that ignores that gets adopted by one department and quietly resented by the other four. So here are the four collisions we see most often, and what each one means for how the software has to be built.
Leakage and operations want more visibility. IT/OT wins by reducing the number of things it has to integrate and defend. Both positions are correct, which is why anything that adds a system gets blocked, and rightly so.
Leakage wants somewhere to work leaks properly. Operations will reject anything that adds a screen without removing noise. The resolution is structural rather than diplomatic: the leak workflow has to feed the operational picture, not sit next to it.
The control room thinks in alarm windows. Asset management plans over decades and needs an audit trail. Same events, opposite clock speeds. So every incident has to be useful twice: once tonight, and once in a capital review a decade from now.
Each specialist wants depth in their own domain. From the top of the building, five separate depths look like five separate pilots, and pilots that never scaled are why the funding conversation is hard. You do not have to commit the whole utility to find out.
What we found, on which network, and what it cost to find it. Roughly monthly, and never a figure we cannot show you the working for.
HULO’s project Lekker (tegen lekken) is co-financed by the European Union, by SNN and by the Dutch Ministry of Economic Affairs.
