Savings Decay Over Time

No matter how good a water savings program may be when launched, time has a good chance of eroding is benefits as the years pass. The long term decay of a program's savings needs to be considered carefully. This article can help you do that.

Real Decay

Water savings generated by a water conservation program may decay over time either because of behavioral backsliding or because of a gradual loss in effectiveness of newly retrofitted equipment. For example, information provided through a landscape audit may initially spur participants to improve their irrigation system upkeep and scheduling, but over time some program participants may lose interest in complying with what they learned. A weather-based irrigation controller retrofit program may lose some effectiveness over time if participating property owners fail to maintain their on-site weather sensors in good working order, or if some of them quit subscribing to the weather signaling service. Showerhead retrofit programs may lose effectiveness over time if some participants replace ultra-low flow showerheads with illegally obtained high-flow showerheads, and so on. These are all examples of real savings decay. Although incorporating such information into demand forecasts or cost-effectiveness analyses is conceptually straightforward, in practice it is rarely available.

Accounting Decay

In other instances, however, savings decay only has meaning in an accounting sense. Care must be exercised to ensure that the accounting methodology is appropriate for the specific questions being addressed.

Cost Effectiveness Analyses

For example, to assess cost-effectiveness of a program it is necessary to discount future costs and future water savings into present value terms so as to account for the time value of money. Conceptually, this amounts to decaying the value of future water savings, not the physical quantity of water savings generated by a conservation program. If the goal is future water supply planning instead of cost-effectiveness assessment, the quantity of future water savings should not be discounted.

Natural Turnover Models

Issues about savings decay in an accounting sense also occur while modeling the natural turnover of plumbing fixtures and appliances, such as toilets, showerheads, clothes washers, etc. Since all of these fixtures and appliances have a finite life, they eventually get replaced. Plumbing codes and appliance efficiency standards ensure that when they are replaced a significantly more efficient fixture or appliance takes the place of the old discarded equipment. Even if water suppliers implemented no equipment retrofit programs, eventually the entire stock of plumbing fixtures and appliances would become more efficient due to the existence of plumbing codes and appliance standards. Active retrofit programs only aim to accelerate what natural turnover would otherwise accomplish. Since active retrofit programs piggyback on natural turnover, water suppliers often decay savings from active programs to prevent overstating program effectiveness. We work out a case below using toilet retrofits as an example to demonstrate why this may be necessary.

Assume that the average life of a toilet is 25 years and that the probability of its replacement in any given year is constant, which works out to a natural turnover rate of 4%. Now assume 100 toilets are replaced with HETs as a result of a water supplier program in year 1. In the absence of such a program, the plumbing code would also have caused these toilets to be replaced with HETs through natural turnover. The column labeled “HETs Credited to Code” shows the number of toilets credited to the plumbing code in each year even though the 100 toilets were initially retrofitted due to a water supplier program. The column labeled “HETs Credited to Water Supplier” shows the number of toilets credited to the water supplier’s replacement program in each year. Key aspects of the accounting include the following:

1. Physical water savings do not decay over time. Replacing 100 toilets this year will generate water savings from 100 toilets each year thereafter. Only the division of credit for the savings between the code and the water supplier program changes over time.
2. Credit for water savings assigned to the code increases over time while credit assigned to the water supplier decreases. This is because it is assumed that the non-efficient toilets would eventually have been replaced with HETs in the absence of a water supplier program. By the end of 10 years approximately 31 toilets would have been replaced by HETs as a result of natural replacement. Thus, plumbing codes receive credit for approximately 31 toilet replacements in year 10 for every 100 toilets retrofitted in year 1 due to water supplier programs. The residual 69 toilets remains credited to the water supplier program in year 10.
3. The credit accounting impacts the calculation of program cost-effectiveness from the perspective of the water supplier but it does not impact the calculation of physical water savings resulting from the replacement of 100 toilets.

Table 1 How 100 HETs Retrofitted by a Water Supplier in Year 1 are Allocated to Code over Time

Year HETs Credited to Code HETs Credited to Water Supplier
1 0 100
2 4 96
3 8 92
4 12 88
~ ~ ~
10 31 69

Note that Table 1 only accounts for water savings associated with water supplier replacement programs. Total water savings from conversion of toilets to HETs is the sum of conversions realized through both natural turnover and water supplier programs. The share of total savings coming from water supplier programs versus natural turnover depends on the scale of supplier programs relative to natural replacement.

Table 2 Relative Contribution of Natural Turnover and Active Conservation

Year Non-HET Stock at Beginning of Year HETs Retrofitted Due to Natural Turnover During the Year HETs Retrofitted Due to Supplier Programs During the Year Supplier Induced Retrofits Credited to Code in Year 10
1 10000 400 100 31
2 9500 380 100 28
3 9020 361 100 25
4 8559 342 100 22
5 8117 325 100 18
6 7692 308 100 15
7 7284 291 100 12
8 6893 276 100 8
9 6517 261 100 4
10 6157 246 100 0
Total 3190 1000 162

Suppose there are a total of 10,000 non-HET toilets at the beginning of year 1, and an agency retrofits 100 HETs per year through an active conservation program. At the end of year 1, 400 toilets would have turned over due to natural replacement, and another 100 due to the water supplier program (Table 2). By year 10, a total of 3190 HETs would have been installed due to natural turnover and another 1000 due to the active program.

A straightforward accounting may conclude that 23.9% (1000 ÷ (3190+1000) of water saved by HET retrofits should be counted towards active conservation programs. But as discussed earlier, by year ten 31 HETs of the 100 retrofitted in year 1 would have been captured through natural turnover, 28 of the 100 retrofitted in year 2, and so on. Of the total 1000 ULFTs retrofitted via active conservation programs, 162 would have been captured via natural turnover. When this fact is accounted for, the contribution of the active conservation program drops to 20% ((1000-162) ÷ (3190+1000)). This then is an example where savings from active programs must be decayed to properly assess the cost-effectiveness of an active retrofit program. For water supply planning purposes, however, the supplier would reduce its demand in year 10 by the savings impact of 4190 HET retrofits, unless one complicates the analysis some more by introducing the notion of “free-ridership.”

Free Ridership

Free riding refers to customers taking advantage of rebates or vouchers to finance equipment retrofits that they would have undertaken anyway even in the absence of such rebate programs. If in Table 2 we assume, for the sake of argument, that every participant in a supplier’s toilet rebate program was a free rider, then the 1000 HET retrofits financed by rebates really are subsumed in the 3190 retrofits caused by natural turnover. Under a 100% free-ridership scenario, program effectiveness would go to zero, while for supply planning purposes the water supplier would factor in savings only from 3190 HET retrofits for estimating its year 10 demand. One can easily factor in any assumption about free-ridership, from zero to 100% depending on the best information available[1]. Free ridership then is potentially yet another reason for decaying savings projected from an active conservation program.

In summary, savings decay may be a real phenomenon or it may be necessitated by the needs of proper accounting. Water conservation professionals must conceptually be clear as to why they are decaying savings, and only then select an appropriate savings estimation methodology.


  1. Whitcomb, J., Free Riders in ULFT Programs, a report prepared for the California Urban Water Conservation Council, 2002.
Has general subjectBenefits and costs +
Has introductionNo matter how good a water savings program may be when launched, time has a good chance of eroding is benefits as the years pass. The long term decay of a program's savings needs to be considered carefully. This article can help you do that. +