First, meet the pile of receipts
The dataset is full of everyday water work: mains, tanks, pump stations, treatment upgrades, PFAS cleanup, meters, and lead-service-line projects. It is not all shiny ribbon-cutting material, which is exactly why it is useful. Use the filters to poke around.
Observed project distribution
Interpretation
Bigger population bands show lower per-capita costs because the rows are usually partial projects, not total system replacement.
Build-a-water-system
Pick a starting shape, then pull the boring-but-decisive levers: people, demand, pipes, treatment, and O&M. The answer moves quickly, because water is heavy and pipes are rude.
Capital components
Not all water is the same water
A clean nearby aquifer is a gift. Seawater is a science project with a power bill. Old pipes are archaeology with invoices. Same 100,000 people, very different checkbooks.
100k-person scenario model
Capital and monthly bill proxyJackson is where the spreadsheet gets interesting
Jackson is useful because it forces a distinction: needing money after a breakdown is not the same as proving the system was simply short on ordinary resources before the breakdown. The numbers let that question breathe a little.
Back in 2003
$43.0M water/sewer revenue against $35.7M operating expense. Boring, in the good way.
By 2021
$33.1M collected water/sewer revenue in 2021 against $60.8M operating expense.
The awkward bit
$131.7M gross receivables and $94.2M allowance for uncollectible bills.