Rainwater Harvesting for Water Supply – By The Numbers

In “Zero Net Water” the case was made for centering water supply on building-scale rainwater harvesting (RWH). Here we look in more detail into the potential of that strategy to provide water supply in Central Texas, parts of which are forecast to have considerable population growth over the next few decades. Since it is in new development where the Zero Net Water concept would be best applied, this area is a prime target for that strategy.

As reviewed in “Zero Net Water”, a modeling process was used to determine the “right-size” of a rainwater harvesting system to supply interior usage in houses. Modeling was executed presuming a 4-person occupancy in “standard” subdivisions and a 2-person occupancy in subdivisions targeted at seniors. A “right-sized” system is one that has a roofprint and cistern volume relative to the expected water demand profile such that backup supply would only be required in the worst drought years, and even then would be rather limited. This is specified so that the demand for backup supply in these houses from our “normal” supply sources would be minimized, and in recognition that a trucked-in backup supply – expected to be the dominant mode of providing that supply for a number of reasons that are not belabored here – would be stressed if backup supply requirements were not so limited.

First we examine locations tributary to the Highland Lakes, which currently provide the water supply for Austin and much of the area around the lakes, including such fast-growing places as Bee Cave and Dripping Springs. The inherently greater efficiency of building-scale RWH vs. watershed-scale RWH noted in “Zero Net Water” is illustrated by modeling these locations in that tributary area:  Brownwood, Burnet, Fredericksburg, Llano, Menard, San Saba and Spicewood. Only in Brownwood and Menard, located further to the north and west in this area, does the modeling indicate that any backup supply would have been required after the extreme drought year of 2011, while the “right-sized” RWH systems would have provided all the interior water supply since then in all the other locations. This contrasts to how the lakes have “performed” as the watershed-scale “cistern” over that period, as they remain chronically low, not “recovering” after 2011 in the way the “right-sized” building-scale RWH systems would have.

The “right-sized” building-scale RWH systems would have provided 95-98% of the interior demands over the recent drought period at these locations. Using building-scale RWH for interior water supply would have relieved the lakes of having to provide that supply, thus they would have been drawn down more slowly if that had been a broadscale practice. So even though backup supplies to provide the 2-5% deficit may have been drawn out of the lakes – or withdrawn from streams flowing into them – the overall result would have been to significantly conserve region-wide water supply over the modeling period.

Now looking at Austin proper, and at Dripping Springs, as representative of the high-growth areas in this region, we see that a “right-sized” building-scale RWH system would have provided 96-98% of interior demands in the recent drought period through 2013. Indeed, even with 2014 having been very dry well into May, the models show that no backup supply would have been required to date in 2014 as well.

Based on a modeled demand rate of 45 gallons/person/day and an occupancy of 4 persons, “right-sized systems for single-family homes around Austin and Dripping Springs require 4,500 sq. ft. of roofprint and a 35,000-gallon cistern to have provided 97-98% of interior demand through the current drought period. These are fairly large, and would impose significant costs, so the impact of better demand control – water conservation – was also examined.

A demand rate of 45 gallons/person/day is reported by the American Water Works Association to be routinely expected for a residence equipped with state-of-the-art fixtures in which the users give “reasonably” conscientious attention to demand control – e.g., it presumes minimal leakage losses, “reasonable” showering time, etc. It is understood, however, that better demand control is readily attainable. My personal experience is a case in point. According to our winter water bills my wife and I have an average interior demand rate of 37 gallons/person/day for our two-person household. As we are served by the watershed-scale Austin Water RWH system, not a building-scale RWH system, we have no particular impetus to “highly” conserve, as would a rainwater harvester who could see the cistern volume dwindling when rain is scarce. The only “highly” efficient appliance in our house is a front-loading washing machine; all the rest are 1990s-era fixtures. One can conclude, therefore, that something in the range of 35-40 gallons/person/day is a demand rate that is readily attainable without any “crimping” of lifestyle.

Indeed, a lower demand rate is typically presumed by those who design and install building-scale RWH systems, with 35 gallons/person/day being routinely presumed. So the models were also run using a demand rate of 40 and 35 gallons/person/day. At 40, a “right-sized” system that would have attained that same 97-98% coverage of interior water demand requires 4,000 sq. ft. of roofprint and a 30,000-gallon cistern. At 35 gallons/person/day, 96-97% of interior demand would have been covered with a 3,500 sq. ft. roofprint and a 25,000-gallon cistern. All these results presume 4-person occupancy in the house, which is above what demographics indicates is the average household size in most single-family residential developments around Austin and in the Hill Country, so it is expected that this sizing criteria would adequately supply the demands in most new houses.

These findings indicate that attaining very good demand control can significantly decrease the scale of facilities needed to “right-size” the building-scale RHW system, which would significantly reduce their costs. A single-story house plus garage and a “normal” area of covered porches/patios might provide 3,500 sq. ft. of roofprint, so an RWH house “right-sized” for a demand rate of 35 gallons would not require “extra” roofprint to be fit into the plan, so would not entail a cost increase to provide the required roofprint. And with the cistern being the costliest component of a building-scale RWH system, reducing its size contributes significantly to rendering the overall system more cost efficient.

With the baby boomers coming to retirement age, and single people and “DINKS” (dual income, no kids) being significant demographics, many building-scale RWH systems may be sized to serve 2-person households, for which the “right-sized” systems would be much smaller. Modeling in Austin and Dripping Springs shows that, with a demand rate of 45 gallons/person/day, a roofprint of 2,500 sq. ft. and a cistern volume of 17,500 gallons would have covered 97-98% of interior demands through the recent drought period. At a demand rate of 40 gallons/person/day, this result would have been attained with a roofprint of only 2,000 sq. ft. along with that 17,500-gallon cistern. If demand rate averaged 35 gallons/person/day, then a roofprint of 2,000 sq. ft. along with a 12,500-gallon cistern would have covered 97-98% of total interior demand. A small single-story house plus garage or carport and a “reasonable” area of covered porch/patio would provide that 2,000 sq. ft. roofprint, thus requiring no “extra” roofprint to be paid for. So, with significantly smaller cisterns being required, this market could more cost efficiently employ a building-scale RWH water supply strategy.

A model was also run covering the drought of record period from the late 1940s to the mid-late 1950s. The worst portion of that drought was from 1950 to 1956. Model results show that for all the scenarios reported above, a “right-sized” building-scale RWH system would have covered 92-95% of the interior water demands through that period. Comparing the rainfall deficits relative to long-term averages, it is seen that the 1950-1956 period was somewhat more “intense” than the recent drought period; while 2011 was the worst year on record, overall the current drought has not (yet) approached the severity of the drought of record. Even under the drought of record condition, however, it is seen that a “right-sized” building-scale RWH system would have provided the vast majority of interior water demands.

Many commercial and institutional buildings would also have a roofprint to water demand ratio that would be favorable to building-scale RWH. For example, a system for a two-story office building in which water usage rate is 5 gallons/person/day (typical toilet and lavatory use by an office employee) might have provided ~99% of water demand through the recent drought period. Whole campuses of such buildings might be built without having to install any conventional water and wastewater infrastructure, using wastewater treated at the building scale, perhaps supplemented by condensate capture, to supply toilets and all irrigation of the grounds, so allowing a smaller cistern to be installed, or allowing a higher water usage rate – e.g., to also cover food service – while still providing essentially all the demand. Capturing roof runoff in the RWH system would also reduce the stormwater management problem in such a development, enhancing the benefit of this strategy.

We can see therefore that building-scale RWH has great potential for relieving stress on the watershed-scale RWH systems that compose our “normal” water supply strategies, and could blunt the need for such high-cost options as desalination, direct potable reuse, or long-distance transfers from remote water sources. So even though building-scale RWH is relatively expensive in capital costs, it may be cost efficient relative to other options, while also offering low long-term operating costs.

One of those costs is for energy to pump and treat water. Building-scale RWH is a strategy that would entail relatively low energy use. Since the water loop is “tight”, water would be pumped only very short distances with little elevation head to overcome. This would save even more water, since it takes water to produce electricity to drive pumps – the so-called “water-energy nexus”.

On the basis of water usage efficiency, then, the building-scale rainwater harvesting strategy is well-worth serious consideration as a major means of serving the increasing demands which would be imparted by the projected growth in Central Texas. The same can be demonstrated for other high-growth regions in Texas, such as the Dallas-Forth Worth area.

Yet the present State Water Plan utterly rejects building-scale RWH as having any merit as a water supply strategy. I am told the reason for this is because the mental model of our controlling institutions sees building-scale RWH as “unreliable” because the cisterns may run dry during severe drought and require those minor fractions of total supply to be added to them from other sources. The counter to this is to think of it as “conjunctive management” of the total water resource, with the RWH systems diverting demand from other sources, decreasing their routine drawdown so that they have the capacity to provide the backup supply.

This highlights that, as noted in “Zero Net Water” there are challenges to be addressed, but those challenges may be less problematic than those posed by desalination, direct potable reuse or long-distance transfer schemes. So water policy makers should be called upon to recognize this clear potential and to incorporate this strategy into their water planning going forward.

It is noted in closing that the analyses reported in this post addressed only interior water usage. As reviewed in “Zero Net Water”, that concept envisions exterior usage – irrigation – to be largely supplied by localized reclamation and reuse of the “waste” water produced in the buildings being supplied by building-scale rainwater harvesting. In itself that tight-looped “decentralized concept” of wastewater management is a more highly efficient strategy – in regard to both money and water – than the conventional long-looped “regional” system, as was generally reviewed in “It’s the infrastructure, stupid”. That aspect of the Zero Net Water concept will be further considered in a future post.


Explore posts in the same categories: Uncategorized

2 Comments on “Rainwater Harvesting for Water Supply – By The Numbers”

  1. bshone59 Says:

    I’m the above-mentioned wife. Ten years ago this month I married the most brilliant P.E. in the field of sustainable water use/reuse.
    My husband’s integrity and vision, with his tenacity for ‘doing the right thing’ must become the ‘norm’ – sooner rather than later.
    The ideas in “Zero Net Water” should be widely implemented if we expect dwindling water resources to sustain us and future generations.

  2. Paul Lawrence Says:

    As always, you’ve given us a new perspective to view the role rainwater should play in our water-challenged world. Like I always say, ‘If it doesn’t work on paper, it most likely, will not work in the real world.’ And this works in both!

Leave a Reply to bshone59 Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: