RUBS vs Submetering ROI: A Worked Breakeven Model
Submetering cut water use 15.3 percent in the national study and RUBS cut none. On water savings alone, a $300 meter pays back in 2 to 6 years at public rates.

Ratio utility billing (RUBS) and submetering both move the water bill from the owner to the resident. Only submetering has been shown to reduce how much water the building uses. This report turns that difference into a breakeven model you can run with your own rates and quotes.
The consumption effect comes from the 2004 National Multiple Family Submetering and Allocation Billing Program Study. Prices come from a 2023 Pacific Northwest National Laboratory rate report prepared for the US Department of Energy. Every other input is a labelled assumption.
Key findings
- Submetered properties used 15.3 percent less water than properties with water included in rent, a saving of 7.96 thousand gallons per unit per year. The 95 percent confidence interval was plus or minus 9.3 percentage points. Source: Mayer and others, 2004, Table ES.1.3.
- RUBS properties showed a difference of plus 0.7 percent, which was not statistically different from zero. Source: same study, Table ES.1.4.
- The study's own owner analysis, at a combined water and wastewater charge of $5.27 per thousand gallons, found benefit-cost ratios of 1.9 to 5.1 for submetering and 4.9 to 7.6 for RUBS. From the owner's cash position, RUBS scored higher because it has almost no upfront cost.
- Regional average volumetric rates in 2021 ranged from $5.86 per thousand gallons for water plus wastewater in the Mountain West to $17.53 on the Pacific coast. Source: PNNL-34006, 2023.
- At those rates, 7.96 thousand gallons is worth $47 to $140 per unit per year. A meter costing $300 per unit pays back from water savings in 2.2 to 6.4 years.
What did the national study measure?
The study was written by Peter Mayer, Erin Towler and William DeOreo of Aquacraft with several co-authors, and funded by the US EPA, apartment industry associations and water utilities. It compared indoor water use at multifamily properties in 13 utility service areas, including Denver, Seattle, San Diego, Phoenix, Tucson, Austin, San Antonio, Portland, Las Vegas and Indianapolis. It used utility billing records for 2001 and 2002, surveys of owners and managers, and site visits.
| Billing method | Annual indoor use per unit (thousand gallons) | Difference from in-rent | Significant at 95%? | Properties in model |
|---|---|---|---|---|
| Water included in rent | 52.19 | Baseline | Baseline | 703 |
| Submetered | 44.23 | -15.3% (plus or minus 9.3) | Yes | 100 |
| RUBS | 52.58 | +0.7% (plus or minus 7.4) | No | 150 |
Source: National Multiple Family Submetering and Allocation Billing Program Study, 2004, Tables ES.1.2 to ES.1.4, statistical model 2.
What are the study's limits?
- The data is more than 20 years old. Fixtures, appliances and meter technology have changed.
- The 13 sites were weighted toward the West and Southwest. The authors say the properties were not selected to represent the whole country.
- The model explained about 25 percent of the variation in water use between properties. Not every submetered property used less.
- Site visits found that 3 of 20 properties reported as submetered were metering hot water only, so the submetered group probably includes some hybrids.
- Across the study's seven analyses the submetering effect ranged from 11 to 26 percent. The 15.3 percent figure is the authors' best estimate, not a guarantee.
The study also found that properties built before 1995 used 10.8 thousand gallons more per unit per year than newer ones, which the authors attribute largely to plumbing fixtures. That is a bigger effect than submetering, and the authors recommend fixture upgrades alongside any billing change.
Who keeps the savings?
This is the question most ROI claims skip. Under either method the resident pays the water bill, so the owner's income statement improves by the amount recovered. The study treated that recovery as the owner's benefit and assumed rent was not reduced.
The conservation saving is a separate matter. When a submetered resident uses less, the resident's bill falls. The owner captures conservation value only where the owner still carries part of the cost. Three common cases:
- The property moves from water included in rent to submetering, and the owner keeps part of the bill, such as common areas or a capped charge.
- Local rules limit what can be passed through, so a smaller master bill reduces the unrecovered remainder.
- The owner competes on total cost of occupancy, so a lower resident bill supports rent or retention. That is a judgment, and this model puts no number on it.
For that reason the model below reports two results: a property-level payback that counts all water saved, and the owner's recovery, which is the same for both methods before fees and rules.
What is the breakeven formula?
Property-level payback on the meters, in years:
Payback = C / (S x P - M)
Cis the installed meter cost per unit.Sis water saved per unit per year, in thousand gallons.Pis the combined volumetric water and wastewater rate per thousand gallons.Mis any annual cost per unit that submetering adds over RUBS, such as meter reading or maintenance.
Rearranged, the breakeven rate for a target payback of N years is:
P = (C / N + M) / S
What do the inputs look like?
| Input | Value | Basis |
|---|---|---|
| Water saved, S | 7.96 thousand gallons per unit per year | Measured. 2004 study, model 2 |
| Rate, low case | $5.86 per thousand gallons ($2.74 water plus $3.12 wastewater) | Published. Mountain West regional average, 2021 dollars, PNNL-34006 |
| Rate, high case | $17.53 per thousand gallons ($5.63 water plus $11.90 wastewater) | Published. Pacific regional average, 2021 dollars, PNNL-34006 |
| Rate, historical reference | $5.27 per thousand gallons | Published. Average of the 13 study sites, 2001 to 2002 |
| Meter cost, C, case A | $300 per unit | Published but old. Retrofit average cited in the 2004 study from 1999 and 2001 literature. New construction was cited at $200 |
| Meter cost, C, case B | $600 per unit | Assumption. Twice the 2004 figure, to test sensitivity. Replace with your quote |
| Added annual cost, M | $0 | Assumption. Both methods are assumed to carry the same billing service fee |
| Equipment life | 10 years before transmitter replacement | Published. The 2004 study assumed replacement at 10 years in a 20-year analysis |
Two cautions on prices. The PNNL figures are volumetric charges for a large industrial meter class, taken from American Water Works Association surveys, and are in 2021 dollars. They are a public proxy for the marginal price of water. They are not multifamily tariffs. PNNL also found that rates rose faster than inflation: an average of 3.0 percent a year for water and 3.2 percent for wastewater in real terms across the utilities it tracked between 2008 and 2021. Current rates are likely higher than the table shows, which shortens payback.
We found no public, neutral benchmark for today's installed submeter cost. That is why case B is an assumption and not a citation.
Worked example: a 200-unit property
Assume a 200-unit property built before 1995, currently on RUBS, with one accessible supply line per unit.
Step 1. Water saved. 200 units x 7.96 thousand gallons = 1,592 thousand gallons a year.
Step 2. Value of water saved. At $5.86, that is $9,329 a year, or $46.65 per unit. At $17.53, it is $27,908 a year, or $139.54 per unit.
Step 3. Capital cost. At $300 per unit, $60,000. At $600 per unit, $120,000.
Step 4. Payback.
| Scenario | Annual value of water saved | Capital cost | Property-level payback |
|---|---|---|---|
| Low rate, $300 meter | $9,329 | $60,000 | 6.4 years |
| Low rate, $600 meter | $9,329 | $120,000 | 12.9 years |
| High rate, $300 meter | $27,908 | $60,000 | 2.2 years |
| High rate, $600 meter | $27,908 | $120,000 | 4.3 years |
Step 5. Breakeven rate. For the meters to pay back within a 10-year equipment life, the combined rate must be at least $3.77 per thousand gallons at $300 per unit, and $7.54 at $600 per unit.
How to read the result
At Pacific coast rates the case is strong at either cost. At Mountain West rates a $300 meter pays back inside its life, and a $600 meter does not, on water savings alone.
The 12.9-year line is the important one. It shows that in a low-rate market the conservation saving cannot carry an expensive retrofit by itself. The decision then rests on other things: a regulation that requires meters, resident disputes, or a sale where the buyer values metered billing.
Remember who keeps the saving. In a full pass-through program the $9,329 to $27,908 lowers resident bills. It reaches the owner only through the three routes described above.
What does the model leave out?
- Collection. In the 2004 manager survey, reported non-payment averaged 6.1 percent of resident water bills at submetered properties and 6.4 percent at RUBS properties. The methods were close, so the model ignores the difference.
- Service fees. Billing companies in the 2004 survey charged an average of $3.29 per unit per month, with a range of $1.50 to $6.14. The model assumes the fee is the same for both methods. Check your state before passing a fee to residents. Our companion report on submetering rules lists states that cap or ban it.
- Complaints. When properties converted, 61 percent of RUBS managers reported resident complaints, against 22 percent of submetered managers. Staff time is real, and the study did not price it.
- Behaviour change over time. The study is a cross-section. It does not show whether savings grow or fade.
- Price response. The study estimated a price elasticity of about -0.27 for multifamily water use measured against utility rates. Savings may be larger where rates are higher. The model holds the saving constant.
- Hot water energy. Lower hot water use also lowers gas or electric cost. The study did not measure it, so the model does not count it.
- Regulation. Some states require meters in new buildings and some restrict RUBS. Where meters are mandatory, payback is not the question.
When does each method win?
| Situation | Likely choice | Why |
|---|---|---|
| Combined rate under about $4 per thousand gallons | RUBS | Even a $300 meter does not pay back within equipment life on water savings |
| Combined rate over about $8 per thousand gallons, one supply line per unit | Submetering | Payback inside 10 years even at $600 per unit |
| New construction | Submetering | Lowest installed cost, and required in some states |
| Trash, stormwater, common-area utilities | RUBS | There is no unit-level meter to read |
| Frequent fairness disputes | Submetering | The bill follows the resident's own use |
The thresholds in the first two rows are the breakeven rates from step 5, rounded.
Method and sources
The model uses one measured effect, two published price points and two cost cases. All arithmetic is shown above so it can be repeated. Figures from the 2004 study are in the dollars of that period and have not been adjusted for inflation. Nothing in this report comes from AirBills customer data.
- Mayer, Towler, DeOreo and others, National Multiple Family Submetering and Allocation Billing Program Study, Aquacraft and East Bay Municipal Utility District, 2004. Full report hosted by East Bay Municipal Utility District. Executive summary and chapter 6 used for water use, costs, fees, non-payment and elasticity.
- Unger and others, Water and Wastewater Annual Price Escalation Rates for Selected Cities Across the United States: 2023 Edition, PNNL-34006, Pacific Northwest National Laboratory for the US Department of Energy Federal Energy Management Program, 2023. Regional average rates and real escalation rates.
How AirBills helps
AirBills supports RUBS by square footage, occupancy or rooms, true submeter billing, and hybrids of the two, with rules set per property. Because it captures every master utility invoice, you can put your own rate and usage into the formula above for each property and see where meters pay back. Read the shorter guide, RUBS in 2026, for the operating trade-offs.