The Cost of Algae Control: Building the Budget Case at a Water Utility

A bad bloom season is expensive, and the people running the plant know it. Producing the figure is the harder part, because no accounting system has a line called algae. It arrives instead as a coagulant contract that ran over, a filter cleaning schedule nobody planned, eight million gallons of backwash and flush water that consumed treatment capacity without reaching a customer, and a fortnight of complaint calls in August.

That is a funding problem rather than an operational one. The operations manager who feels the cost does not own the budget lines it hides in, and the finance director who owns those lines sees numbers that moved for reasons nobody has separated out.

This is a method for separating them, built for the September to December window when the request actually gets written. It covers the seven line items to pull, how to make the comparison hold up under scrutiny, what four documented sites reported, and how to check any supplier’s savings percentage before it reaches a board paper.

Why algae never appears in the budget as algae

Algae rarely carries a cost code of its own. The spend is real, and it is distributed. It sits in the coagulant contract, the polymer contract, the additional powdered activated carbon ordered during difficult taste-and-odour periods, the electricity that runs the backwash pumps, the overtime that covers a filter cleaning schedule nobody planned for, the laboratory invoices, and the taste and odour calls in August.

This is not a new observation. Dunlap, Sklenar and Blake examined the same problem in Journal AWWA under the title A Costly Endeavor: Addressing Algae Problems in a Water Supply. All three authors were at a consultancy rather than a treatment supplier.

That distribution is why the cost of algae control is hard to fund. A capital request has to name a saving, and every saving here belongs to somebody else’s cost centre. The operations manager who feels the problem does not own the chemical budget. The finance director who owns the chemical budget sees a number that moved and attributes it to unit price.

What resolves it is not a better argument. It is an audit of six or seven line items you already track, over two comparable seasons, with the confounders written down. This article covers which line items to pull, how to make the comparison defensible, what four documented sites actually reported, and how to check any supplier’s savings percentage before it reaches your board paper.

For the qualitative picture of where those savings arise across the treatment chain, our earlier note on optimising water treatment works processes maps them stage by stage. This article is about putting numbers on them.

The seven line items where algae actually shows up

Each of these already exists in your accounting system. None of them is labelled algae.

1. Coagulant and polymer dose

Algal biomass raises turbidity and organic load, so coagulant and polymer demand rises with it. This is often the largest single number and among the easiest to extract, because dosing records are kept daily and unit prices are contractual. Pull kilograms or litres per day alongside raw water turbidity for the same dates.

2. Algaecide and in-reservoir chemical spend

Algaecide spend is already a discrete line, which makes it the cleanest baseline available. Where in-reservoir chemical treatment is eliminated rather than reduced, the line can disappear entirely rather than simply shrink.

3. Backwash and run-to-waste volume

Water used to wash filters and flush clarifiers has consumed treatment capacity without reaching a customer. At plants that draw washwater from clearwater storage it is treated water directly; elsewhere the cost is the capacity and energy it displaced. Either way it is real and rarely costed. Most plants log backwash volume already but few cost it, because it does not arrive as an invoice. Multiply the volume by your own cost per unit produced, not by a tariff.

4. Filter run length and filter availability

Shorter runs mean more backwashes, more units in the wash queue and less clearwater production per day. Unit filter run volume, documented at treatment plants, is the metric that captures this properly, and plants that track it have the cleanest evidence available.

5. Staff hours

Unplanned filter cleaning, manual clarifier work, extra sampling runs and out-of-hours callouts. Timesheets rarely code to algae, so this usually has to be reconstructed from a supervisor’s recollection of how many shifts were consumed. Reconstructed numbers are weaker evidence, so treat them as a range rather than a figure.

6. Laboratory and monitoring

Additional taste and odour panels, cyanotoxin analysis during an event, and increased sampling frequency under an advisory. Commercial laboratory analysis of a full permit suite is expensive enough that a single bad season is visible in the invoices.

7. Customer contacts and taste-and-odour response

Geosmin and MIB are detectable by people at concentrations far below anything that matters to health, which means a compliant plant can generate a complaint spike. The cost is call handling, communications, and in the worst case a public advisory. Where contact reasons are coded, this is retrievable, and the wider economic impact of blooms is documented separately. Where they are not coded, the category still exists but has to be estimated.

Energy sits across several of these rather than standing alone: backwash pumping, mixers and aerators, and any process brought online only during a bloom.

Making the cost of algae control defensible

Where the records allow, compare at least two bloom seasons matched month for month. Annual totals hide the thing you are trying to measure, because a bloom is a three or four month event and an annual average dilutes it into nothing.

Four rules make the difference between a number your board accepts and a number your engineer can pull apart.

1. Match the months, not the years. July to October against July to October. If your bloom window shifted, say so and show both windows.
2. Separate volume from price. Chemical spend can fall because you used less or because the contract renewed cheaper. Report quantity and unit price as separate columns, always.
3. Write down what else changed. A new polymer, a pH adjustment, a filter media replacement, a different abstraction depth, a wetter spring. Every one of these confounds the comparison, and naming them is what makes the rest of the number credible.
4. Normalise for production. Express chemical use per megalitre or per million gallons treated, not as a total, or a demand change will read as a saving.

One point sits above the other four. Establish the baseline before you install anything. The financial baseline comes from chemical records, production figures, SCADA, filter logs, laboratory invoices and timesheets. Water quality monitoring is a separate matter: it does not create the financial baseline, it helps explain afterwards why the numbers moved. A baseline reconstructed afterwards from memory and invoices is worth a fraction of one measured on purpose, and by then you cannot go back and get it.

What four documented sites reported

These are figures the operators themselves put on the record. They are reported outcomes, not the output of the method above, and the categories behind each one differ. Two are US drinking water systems, and two are irrigation and recreational reservoirs at the same district, included because their prior chemical spend is documented and the payback contrast is instructive.

Drinking water evidence: Donner Summit Public Utility District, California

Lake Angela is an 18-acre drinking water reservoir with a bloom history going back to 2009. Backwash and clarifier flush volume across July to October fell from 17.78 million gallons in 2024 to 9.77 million gallons in 2025, a reduction of 8.01 million gallons or 45 percent. Influent turbidity to the treatment plant fell 67 percent on average across the same months, from 2.8 to 0.9 NTU.

The district reported saving 86,000 dollars in the first year from reduced backwash and flush water alone. Steven Palmer, the district’s general manager, noted that the figure was close to what they had estimated before purchase, and that it excludes chemical costs because it was too early to quantify those. The board approved a further two years of treatment in November.

Two details matter more than the percentages. The saving was forecast before the purchase and the first-year result came in close to that estimate. And the customer named a category he had deliberately left out, which is the single most useful sentence in the whole case.

Drinking water evidence: American Water, Canoe Brook, New Jersey

The only one of these documented in the sector literature by utility personnel rather than in supplier material. Schneider, Weinrich and Brezinski published the results in Journal AWWA, volume 107 issue 10, pages E533 to E542, and all three authors were American Water or New Jersey American Water staff reporting on their own utility’s water. A separate AWWA Opflow case report covers a different site. Against the equivalent period in 2013, the plant recorded 127 percent longer filter runs, 83 percent higher unit filter run volumes and 19 percent lower combined filter effluent turbidity. Chemical consumption at the plant fell by 22 percent, and algaecide use in the raw water reservoir was eliminated rather than reduced.

Reported OPEX savings were approximately 87,800 dollars, calculated from projected chemical savings against 2013, elimination of copper treatment in the reservoir, and reduced monitoring costs. Return on investment was 1.8 years. The project received a Business Achievement Award from Environmental Business Journal.

Supporting reservoir evidence: Mahr Reservoir and South Lake, Vallecitos Water District

Neither is a drinking water reservoir, which is why they sit here rather than above. They are included for one reason: the prior chemical spend is on the record for both, and the payback difference between them is the most useful thing in this article.

Copper sulfate spend at Mahr Reservoir peaked at 44,530 dollars in 2013. At South Lake, a park lake with public access, it peaked at 18,677 dollars in 2017. Both are recorded peak years rather than typical annual spend. Both reservoirs have operated without chemical treatment since 2018. Mahr returned roughly 25,000 dollars a year and reached payback in 2.5 years. South Lake returned 84,354 dollars in total and reached payback in five years.

Payback tracks what you already spend

Compare those two reservoirs at the same district, with the same technology, installed in the same year by the same operator. One paid back in 2.5 years and one took five.

One major difference was the scale of avoidable chemical spend already present. Mahr’s recorded peak annual chemical spend was 44,530 dollars, against 18,677 dollars at South Lake. Those are peak years rather than recurring annual figures, so read them as an indicator of scale rather than as the amount displaced each year. Other things differed too, including basin size, public access and the capital deployed, so this is not a controlled comparison. The general principle still holds: higher existing avoidable operating cost creates more room for a shorter payback, all else being equal. That has a useful consequence for anyone building a budget case.

The payback range can be estimated from your own chemical line before any supplier is involved. Take your worst recent bloom year, add the algaecide spend to the incremental coagulant and polymer, add the backwash and clarifier flush volume at your own cost per unit produced, and set that against a quoted capital cost. Whether the result clears your organisation’s acceptable investment horizon is then a question your own records answer.

Across the three sites with a reported payback figure, the range runs from 1.8 years to five years. A site carrying very high avoidable spend could in principle pay back faster than any of these. Even so, an unusually short quoted payback deserves the same five questions as any other figure, and that applies to our own older material where it has quoted one.

What does an algae control system itself cost?

Anyone building this case needs the other side of the equation. There is no single meaningful benchmark price, because system cost is site-specific, and a handful of variables drive it.

• Surface area and basin geometry. Coverage is designed for the shape of the water body, not scaled from acreage, so a long or heavily embayed reservoir needs more units than its area suggests.
• Number of treatment units and their configuration. Whether monitoring is combined with treatment or deployed separately changes both the equipment count and the data you get.
• Monitoring specification. Which parameters, at what depths, and whether vertical profiling is required.
• Installation and mooring. Access, annual water level range and anchoring requirements.
• Service and data provision. Ongoing programme adjustment, reporting and support, which is an operating cost rather than a capital one.

Which is precisely why the sequence in this article runs the way it does. Establish your avoidable operating cost first, from your own records. Then ask for a site-specific capital figure and test it against that number. A quote assessed against a measured baseline is a decision. A quote assessed against a vendor’s percentage is a guess.

How to audit a savings percentage

Ultrasonic algae control now has several suppliers, and published claims range from around 20 percent chemical reduction to figures approaching 100 percent. Both can be true of different sites. Neither is useful until you know what sits behind it.

Five questions resolve almost any claim.

• Against what baseline? A specific prior year, or an unnamed average. American Water’s 22 percent is measured against 2013 specifically, which is checkable.
• Which months? A bloom-season comparison and a full-year comparison give very different percentages from identical data.
• What is included in the number? Chemicals only, or chemicals plus labour plus water loss. Donner Summit’s 86,000 dollars is water volume alone and explicitly excludes chemicals, which makes it a floor rather than a ceiling.
• Who measured it, and is it published? A supplier press release, a customer statement, or a peer-reviewed paper are three different grades of evidence.
• What else changed that season? If the answer is nothing, the claim has not been examined properly.

Ask the same five of any supplier including us. A percentage that survives them is defensible enough to consider in a board paper, though surviving five questions does not make a weak methodology sound. One that does not survive them should not reach the finance director’s desk.

What a pilot has to produce to be worth anything

Free and subsidised pilots are common in this sector during bloom season, and a free pilot that produces no usable number has cost you a season you cannot get back. Cost is not the variable that matters. Evidentiary design is.

A pilot that can support a capital request needs five things.

• A measured baseline established before installation, covering at least one full bloom window.
• An observation period that matches the baseline months exactly.
• An untreated comparison where the hydraulics allow it, whether a second basin, an adjacent reservoir or an upstream point.
• Instrumentation that has been calibrated against laboratory samples, with outage periods excluded from any day counts.
• An agreement in writing, before installation, about which cost categories will be counted and who calculates them.

The fifth is worth settling before the rest. Most pilot disputes turn out to be about which costs were in scope rather than about whether the water improved.

Building the request

Boards and councils tend to approve one page rather than thirty. The structure that works follows the same order as the audit.

1. The operational failure, stated in the plant’s own terms and units.
2. What the worst recent season cost, broken into the line items above, with the confounders listed.
3. What the intervention is expected to change, and in which of those line items.
4. A payback range rather than a figure, with the assumptions that produce each end of it.
5. What is not included, and why.
6. How the result will be measured, agreed before approval rather than after.

Point five carries more weight than it looks. Naming the exclusions is what separates analysis from advocacy, and finance directors have read a great many pieces of advocacy.

What this method does not do

Being clear about the boundaries is what makes the rest usable.

• It does not produce a single number. The output is a range with named assumptions, which is both the honest answer and the more defensible one.
• It cannot fully recover a baseline you never measured. Where no pre-intervention season exists, you can reconstruct one from historical operating records and state the limitations of that reconstruction openly, or measure a proper untreated season before making strong financial claims. A treated season cannot serve as the baseline.
• It does not value avoided risk. A cyanotoxin advisory that did not happen has no invoice. That belongs in the narrative section of the paper, not the arithmetic.
• It does not address the nutrient load. Algae control at the reservoir reduces what the plant absorbs. Watershed and nutrient management is the durable fix and runs on a longer timescale, so most working programmes combine them.

Algae control is also not the answer to every turbidity problem. Where suspended solids are predominantly mineral rather than biological, the cost case above will not materialise, and separating the two fractions is a prerequisite rather than a detail.

Frequently asked questions about the cost of algae control

What does an algal bloom actually cost a water utility?

It depends on the plant, and it never appears as one number. The cost is distributed across coagulant and polymer dose, algaecide spend, backwash and run-to-waste volume, filter availability, staff hours, laboratory analysis and customer contacts. Documented sites have reported figures from tens of thousands of dollars a year upward, but those figures cover different categories and are not directly comparable, which is why the method matters more than any published number.

How do I calculate the cost of algae control at my own site?

Pull two comparable bloom seasons where possible, matched month for month, for chemical quantity and unit price separately, backwash volume, filter run length, staff hours and laboratory invoices. Normalise chemical use per unit of water treated. Then write down every other change in the same period, because those confounders are what determine whether the number holds up.

What is a realistic payback period for ultrasonic algae control?

Across the three sites with a reported payback figure, the range runs from 1.8 years to five years. Payback depends heavily on how much avoidable cost a site already carries. Two reservoirs at the same district, with the same technology installed in the same year, reported 2.5 and five years, and one major difference between them was peak annual chemical spend of 44,530 against 18,677 dollars.

Why do two utilities with the same system report different payback?

Largely because of how much cost there was to displace. Mahr Reservoir’s copper sulfate spend peaked at 44,530 dollars and it reported payback in 2.5 years. South Lake peaked at 18,677 dollars and reported five. Same technology, same operator, same installation year, though basin size, public access and capital deployed also differed, so treat it as indicative rather than controlled.

How much can algae control reduce chemical use?

Published figures across the sector vary widely, and both ends can be accurate for different sites. American Water’s Canoe Brook plant recorded a 22 percent reduction in plant chemical consumption measured against 2013 specifically, and eliminated algaecide use in the raw water reservoir entirely. Treat any percentage as unusable until you know the baseline year, the months compared and the categories included.

How much does an algae control system cost to install?

There is no single meaningful benchmark price, because cost is site-specific. It is driven by surface area and basin geometry, the number of treatment units and their configuration, the monitoring specification, installation and mooring requirements, and the ongoing service and data provision. Establish your avoidable operating cost from your own records first, then test a site-specific quote against it.

How do I check a supplier’s savings claim?

Ask five questions: against what baseline year, which months, which cost categories are included, who measured it and is it published, and what else changed that season. A claim that survives all five is defensible enough to consider in a board paper, though surviving them does not make a weak methodology sound. One that does not survive them should not reach the finance director.

Does reduced backwash volume count as a saving?

Yes, and it is the category most often left out. Washwater and clarifier flush consume treatment capacity without reaching a customer, so cost the volume at your own cost per unit produced rather than at a customer tariff. Donner Summit reported 86,000 dollars in the first year from this category alone, explicitly excluding chemicals.

Is a free pilot worth taking?

Only if it is designed to produce a defensible number. That needs a measured baseline before installation, a matched observation period, an untreated comparison where hydraulics allow, calibrated instrumentation, and a written agreement about which cost categories count. Free is not the variable that matters.

What if we never measured a baseline?

Two options remain, and both are weaker than measuring one. Reconstruct a baseline from historical operational records, chemical purchases and filter logs, stating the limitations of that reconstruction openly. Or measure a proper untreated season before committing capital. A treated season cannot stand in as the baseline, because it no longer represents the condition being priced. Measuring one season in advance is the least expensive part of this exercise.

How do I put a number on taste and odour complaints?

Start with contact volume if your utility logs contact reasons, then apply your own cost per contact handled. Add communications time and any advisory costs separately. Geosmin and MIB are detectable well below health-relevant concentrations, so a fully compliant plant can still generate a complaint spike, which is why this category is real even when compliance was never at risk.

Should the capital request name a single savings figure?

No. Name a range with the assumptions that produce each end of it, and name what the range excludes. A request that states its own exclusions reads as analysis. One that claims every category reads as a sales document.

Does this cost case apply if our turbidity is mostly mineral?

No. Where suspended solids are predominantly mineral rather than biological, algae control will not move the numbers above. Establish the organic and inorganic split with laboratory analysis before building any financial case on it.