Archive for August 2014

Water for DFW – Building-scale rainwater harvesting vs. Marvin Nichols

August 7, 2014

In the last post we reviewed the potential of building-scale rainwater harvesting (RWH) as a water supply strategy in the high-growth area around Austin, in Central Texas. Here, we examine its potential in another high-growth area of Texas, the Dallas-Fort Worth area, commonly called the Metroplex. And then we will contrast that strategy with doubling down on the watershed-scale rainwater harvesting strategy, as may be represented by the proposed Marvin Nichols Reservoir.

To gain an appreciation for the potential of building-scale RWH in and around the Metroplex, modeling was executed for the following locations: Athens and Terrell to the east-southeast, Ferris closer in to the south, Cleburne to the southwest, Weatherford to the west, Bowie to the northwest, Sherman to the north-northeast, and Denton closer in to the north-northwest. Ringing the Metroplex, these locations offer an overview of conditions all around it.

As was the case for the modeling results of the Central Texas locations, it was seen that “right-sized” building-scale RWH systems around the Metroplex would have provided 97-99% of total interior supply through the recent drought period for houses modeled with a presumed average water usage rate of 45 gallon/person/day. But around the Metroplex, the “right-sized” systems would be somewhat smaller than would be required around Austin. Recall that the “right-sized” system there to serve a 4-person household would be a roofprint of 4,500 sq. ft. and a cistern volume of 35,000 gallons. In Bowie, Weatherford and Cleburne, the “right-sized” system for a 4-person household would require only 3,750 sq. ft. of roofprint, paired with a 25,000-gallon cistern in Cleburne and Weatherford and a 27,500-gallon cistern in Bowie. All other locations would require 3,250-3,500 sq. ft. of roofprint and 20,000-25,000 gallons of cistern capacity. It is expected that a one-story house plan with a 2-car garage plus a “typical” area of covered patios/porches could provide a roofprint of 3,000-3,500 sq. ft., so these modeling results indicate many houses in/around the Metroplex would not require any “extra” roofprint to be added on.

As reviewed in the last post, a usage rate of 45 gallons/person/day should be readily attainable by most people, given a house fitted with the current stock of water fixtures, but a lower rate could be routinely attained by people even moderately attentive to conserving water. If a usage rate of 40 gallons/person/day were routinely attained around the Metroplex, the “right-sized” systems that would have provided 97-100% of total interior supply for a 4-person household through the recent drought period would require 3,000-3,500 sq. ft. of roofprint and 17,500-20,000 gallons of cistern capacity for a 4-person household.

Just as in Central Texas, with the baby boomers reaching retirement age and demographics tending toward more one and two-person households in all age groups, a significant part of the market might be made up of houses that could be “right-sized” for a 2-person occupancy. Modeling this occupancy around the Metroplex, at a water usage rate of 45 gallons/person/day the “right-sized” system that would have covered 97-99% of total interior demand would have a roofprint of 1,750-2,000 sq. ft. of roofprint and a cistern capacity of 10,000-15,000 gallons. At a water usage rate of 40 gallons/person/day, a “right-sized” system covering 97-100% of interior demand would require a roofprint of 1,750 sq. ft. and a cistern capacity of only 10,000 gallons, except for Bowie where a 12,500-gallon cistern would have been required. Since it is expected that a one-story house plan plus garage or carport and modest area of covered patios/porches would provide about 2,000 sq. ft. of roofprint, this market could use building-scale RWH without requiring any “extra” roofprint, and would incur relatively modest cistern costs.

So the water supply potential of building-scale RWH around the Metroplex is pretty clear. Yet there is not a mention of this strategy in the planning documents of state planning Region C, the area around the Metroplex. Actually there is no respect shown for this strategy in any of the regional plans, and the state water plan explicitly dismisses it, stating, “While it is often a component of municipal water conservation programs, rainwater harvesting was not recommended as a water management strategy to meet needs since … the volume of water may not be available during drought conditions.” Which is to say that because a “right-sized” system may need 1-3% of the total supply from alternative sources during severe drought periods, this strategy is deemed not to exist at all!

This is likely due to the water planners being guided by a mental model that does not comprehend building-scale RWH as a consciously chosen broadscale strategy, as perhaps the water supply strategy in whole developments.  This was the subject of an investigation, funded by the Texas Water Development Board, that I ran a couple years ago, in which it was brought out that this strategy confers a number of advantages relative to conventional – or watershed-scale RWH – water supply systems. One of the issues considered was provision of backup supply, but only on the basis of the “mechanics” of delivering it. Not fettered by the mainstream’s mental model, it had not occurred to me to question the whole strategy because some small amount of backup supply would no doubt be needed – indeed, the whole idea of “right-sizing” was to cover water demands in all but the worst drought periods and plan on providing a backup supply, presuming that the relieved capacity offered by building-scale RWH would make such a supply available from the sources so relieved.

Still, this does beg the question of from exactly where that backup supply would be derived. As noted in the last post, the building-scale RWH strategy should be considered in the context of “conjunctive management”. Building-scale RWH would divert the vast majority of the demand off of the conventional sources, so decreasing the routine drawdown of those supplies, thus leaving in them the capacity to provide the small amount of backup supply. Of course, if it is presumed that any development on building-scale RWH is in addition to rather than in place of development drawing from those conventional supplies, and that this other development would be of such extent that it would tax the available supply sources during those drought periods, then there may indeed be a question of whether the capacity to provide backup supply for building-scale RWH systems would be available. It will require another whole study to examine how a conjunctive management concept could work in practice. Until the mainstream water planners can get around their mental model and recognize the inherent potential of building-scale RWH, however, it is unlikely that any such study would get funded.

Around the Metroplex, however, modeling shows that, unless the drought gets more severe than has been experienced since 2007, essentially 100% of interior demands could be provided by upsizing the roofprint and/or cistern volume only a modest amount above what is reported above. The worst case would be in Bowie, where a roofprint of 4,000 sq. ft. and a cistern capacity of 30,000 gallons would be required for a 4-person household using water at a rate of 45 gallons/person/day.

So we can provide interior water usage with building-scale RWH, but why should we, rather than continuing to expand and perpetuate the watershed-scale RWH strategy? Consideration of the problems and hazards of building Marvin Nichols Reservoir offers some insights into that.

Marvin Nichols Reservoir would be located in northeast Texas, about 115 miles east-northeast of the Metroplex. The Region C report offers this about that project:

“As a major reservoir project, Marvin Nichols Reservoir will have significant environmental impacts. The reservoir would inundate about 68,000 acres. The 1984 U.S. Fish and Wildlife Service Bottomland Hardwood Preservation Program classified some of the land that would be flooded as a Priority 1 bottomland hardwood site, which is “excellent quality bottomlands of high value to key waterfowl species.” … Permitting the project and developing appropriate mitigation for the unavoidable impacts will require years, and it is important that water suppliers start that process well in advance of the need for water from the project. Development of the Marvin Nichols Reservoir will require an interbasin transfer permit to bring the water from the Sulpher River Basin to the Trinity River Basin. The project will include a major water transmission system to bring the new supply to the Metroplex.”

Unstated is that many people in the area that would be impacted are highly opposed to this project, due in large part to those “unavoidable impacts.” This is a battle of economic interests – those in the Metroplex that purport a need for this water vs. those, such as the timber producers, that would be eliminated by the reservoir. Indeed, the official position of the planning process in planning Region D, where the reservoir would be located, is in opposition to the project, and it is not included in their plan. This contrasts with deriving “new” water supply from building-scale RWH, which would have positive economic impacts in Region C – benefiting businesses that would design, install and maintain the building-scale RWH systems – and no negative impacts in Region D.

As noted, utilizing in the Metroplex any of the water collected in this reservoir would require a huge investment in transmission facilities – pipelines and pump stations – and on-going operating costs to maintain them and for energy to run the pumps. Of course the water would need to be treated, also entailing considerable energy requirements. Since it takes water to make energy, this would cut into the water use efficiency from this source. And making that energy would also generate greenhouse gases, which would exacerbate the already problematic impacts of climate change on regional water resources. This contrasts with the building-scale RWH strategy, which would not require any transmission facilities and would require far less energy to treat and pressurize the water for use within the building.

As the Region C report states, it will take a long time to permit and build this reservoir and the transmission facilities, meaning delivery of the first drop of water is decades away. In contrast, the building-scale RWH strategy could begin delivering water supply immediately, and grow in lockstep with demand, one building at a time.

The passage from the Region C report refers only peripherally to the ecosystem services that flooding the land would eliminate or damage, noting only loss of habitat for “key waterfowl species”, without quantifying how critical to the well-being or survival of those species that loss may be. That of course would be sorted out in the process of preparing the environmental impact analysis that will be required as part of the permitting process, another expense that would be obviated by the building-scale RWH strategy. But those ecosystem services go well beyond their impact on birds. Eliminating the timberlands loses the oxygen production and carbon sequestration they provide, along with habitat for many other plants and animals. Forests are also important to maintaining water quality and to the storage and release of water for environmental flows, which would instead need to be provided “artificially”, with water from the reservoir of degraded quality, including thermal impacts. None of these “externalities” figure into the cost of water projected for this strategy, significantly “warping” the analysis.

There would also be significant losses from the watershed-scale rainwater harvesting system this reservoir would create. Huge evaporation losses from the reservoir would be incurred, and there would be significant losses in the transmission system. In contrast, the building-scale RWH strategy would suffer no such losses.

The Region C report also states, “… the unit cost [of the water supply the reservoir would provide] is less than that of most other major water management strategies.” While at the end of the day the overall direct cost of Marvin Nichols Reservoir and its required infrastructure might be less than the aggregate direct cost of the number of building-scale RWH systems that would provide equivalent supply – which it is noted has not been developed in the Region C report for comparison – much of the cost of the former would need to be expended well up front of delivering the first drop of water to the Metroplex, and all that investment would be at risk. The costs of the building-scale RWH strategy, on the other hand, would be incurred incrementally, one building supply system at a time, so the delivery of supply would pretty directly track the capital requirements. This works with the “time value of money” to defray the global long-term cost of the building-scale RWH strategy. So it is not at all clear that the global cost of the Marvin Nichols option, even neglecting the externalities which the Region C report ignores, would be less.

In summary, broadscale implementation of building-scale rainwater harvesting may provide sufficient supply so that the conventional sources would be sufficiently “relieved”, allowing growth to be sustained without requiring new reservoirs. And it may do so at a cost that would be competitive with the global costs of continuing to extend and perpetuate the watershed-scale rainwater harvesting strategy, which would require going far afield to obtain additional new supply. Yet this is, quite consciously, the road not taken by the water planners in Region C. Or, as noted, anywhere else in the state where building new reservoirs, raiding remote aquifers, and other conventional supply strategies are purported to be needed to support projected growth. Time to re-evaluate?