Fourth in a four-part series on the Roanoke Valley’s water supply.
Someday the Roanoke Valley will need more water.
Exactly when is unclear: A study done in 2010 put the date at which the valley would run out of water a half-century into the future: 2060.
Reality has lagged behind. The Roanoke Valley has used a lot less water than projected, mostly because of more water-efficient appliances and because leaky pipes have been replaced.
That run-out date is now further out, perhaps 2100 or so.
One thing that could change that, and pull that date closer, is whether the Google data center complex someday needs more than the 2 million gallons a day it has initially reserved. Google has said that “someday” it may need 8 million gallons a day. (Disclosure: We have received a grant from the Google News Initiative to support our expansion in the New River Valley, but donors have no say in news decisions; see our policy.) Perhaps that day will never come — perhaps technology will be invented that reduces the thirst of data centers. However, caution dictates that we not count on such inventions. That’s the rationale behind the 24-page agreement that Botetourt County and the Western Virginia Water Authority signed in 2025 that spells out each party’s obligation “to identify and develop future water supply sources.”
That raises the obvious question: Where might this water come from?
That 2010 water study conducted for the Roanoke Valley-Alleghany Regional Commission looked at that very question, a decade and a half before Google picked the Roanoke Valley. It identified 11 different options (some with multiple variations). Significantly, none of those involved building a new reservoir.
There was no single preferred option, partly because some options were better-positioned to serve some parts of the Roanoke Valley than others. While the whole region is tied together in a single system, there are still the practical issues of moving water from one place to another underground. For instance, the report said that if the issue were simply providing more water to Bedford and Franklin counties, that would best be solved by drawing (and treating) more water from Smith Mountain Lake. For some other jurisdictions, the simplest solution would be to drill more wells. Many of these options weren’t either/or scenarios. The valley could theoretically pursue multiple options at once. Then there were other options that were deemed better at addressing valley-wide water supply issues.
The different scenarios were evaluated six ways: applicability (does it address the problem?), safe yield (can it supply enough water?), potential environmental impacts, potential human impacts, cost and availability (part of this involved the politics each would encounter).
The option that ranked the highest was drawing water from Smith Mountain Lake and pumping it to Carvins Cove. It drew the highest ratings in five of six categories, the only option to do so.
The report identified two ways to draw water from the lake: either directly or through the Smith Mountain Lake Regional Water Treatment Plant at Moneta. If the latter, the report said the authority could draw 10 million gallons a day (which would cover Google’s potential need for 8 million gallons a day, something the report authors couldn’t have foreseen). If the former, no specific water amount was listed, other than to say “a large amount” is available. At full pond, Smith Mountain Lake holds roughly 749.4 billion gallons of water, although the lake isn’t at full pond right now.
The only category where the lake option didn’t get the highest score was “availability,” which covers the politics of being able to access a particular water source. The report didn’t say why the lake option ranked lower, but it appears to be because of all the permits that would be required. Smith Mountain Lake wasn’t built to look pretty; it was built as an energy source for Appalachian Power, so taking water out of it potentially could have an impact on how much energy could be generated. There are also lots of people living along the lake who aren’t happy with the low water levels now; they might be even less happy at the prospect of water levels going lower — even if the percentage involved is minuscule. Let’s say we needed to take 10 million gallons a day out of the lake for water supply; that’s 0.1% of the lake’s volume.
The option that ranked the highest for availability seems almost too simple: pumping water from Spring Hollow to Carvins Cove and then replenishing Spring Hollow by drawing more water from the Roanoke River. That’s an indirect way of saying we should just draw more water from the Roanoke River.
The Spring Hollow option drew mixed reviews overall.
Spring Hollow tied with the Smith Mountain Lake option as scoring the best under applicability; the report noted that the water treatment plant at Carvins Cove can process far more water than is presently withdrawn from the cove, so pumping in water from Spring Hollow would allow that facility to be used at a higher capacity. (The cove’s treatment plan can handle 28 million gallons per day, but the safe yield at the cove is 18 million gallons per day, so even when there’s a maximum draw from the cove, there’s still 10 million gallons per day of unused treatment capacity.)
However, the Spring Hollow option ranked low under the “safe yield” category “because the Roanoke River is somewhat limited in flow, and it may not always be able to be drawn, particularly in times of drought.” The report did not list how much water might be available under this option or what the cost would be. Keep in mind that this 2010 report was not intended to be the definitive report on options.
The report also looked at two other options that have been talked about over the years: drawing water from the New River or the James River. Both of these options begin with political/legal/environmental challenges — they involve moving water from one water basin to another. Virginia law doesn’t prohibit that, although there have been attempts over the years to do so. Carvins Cove currently involves some interbasin transfer; some water from Catawba Creek (which flows into the James River) is piped via a tunnel into the cove. That tunnel was built in 1974; today, that would require a lot more permitting than was involved then.
The 2010 report did rank the New River and James River as options equal to Smith Mountain Lake in terms of the amount of water availability. However, it said the distance from the New River Valley to the Roanoke Valley would drive up the cost of moving water from the New River and that large-scale interbasin transfers from the New River Valley might prove “problematic environmentally.”
The same interbasin issue presumably would challenge a James River option, too (despite that Catawba Creek precedent). Interestingly, though, the report said piping water from the James River to Carvins Cove might be cheaper than piping it from Smith Mountain Lake.
This is where this 2010 report is confusing. The official rating listed the Smith Mountain Lake option at $108 million in 2010 dollars, but did not put a price tag on the James River option.
However, the report cited a previous report from a different author, in 2003, that said the James River option would be the cheaper of the two. It updated those cost figures to 2009 dollars, and said the James River option would cost $125 million compared to $163 million for the Smith Mountain Lake option. If we run all these figures through the federal government’s Inflation Calculator, the James River option comes out to $198.4 million while the Smith Mountain options run $167 million to $258.71 million, depending on which estimate you prefer.
However, the cheapest option of all is pumping water from Spring Hollow to Carvins Cove; that option was put at $12 million in 2009 dollars, which would make it $19.05 million today. The problem is there’s a lot less water available.
In the end, the main choices seem to come down to Smith Mountain Lake (lots of water but potential permit challenges related to using an energy source) versus James River (lots of water but potential permit challenges related to interbasin transfers) versus Spring Hollow/Roanoke River (less water available but much cheaper and more doable).
All these options would involve building pipelines. The region just went through the prolonged controversy of the Mountain Valley Pipeline. That was a natural gas pipeline; would there be less controversy over a water pipeline? I’m not an engineer, but I’ll just point out that the Smith Mountain Lake pipeline would have to make its way through or around a more populated area (the Roanoke metro area). A Spring Hollow-to-Carvins Cove pipeline could basically follow Interstate 81, a less populated part of the valley. The James River pipeline would be entirely within one locality — Botetourt County. (Fun fact: Most of Carvins Cove is in Botetourt.)
Those were the options in 2010, and they’re not likely to be any different at some point in the future, although the rankings might change. My main takeaway from reading this report is that if we’re thinking about future water supply for the Roanoke Valley, we’re not talking about flooding some place to build another Carvins Cove or Spring Hollow. However, we would be talking about building pipelines — the question is from where.
Want more political news and analysis? Sign up for West of the Capital, our weekly political newsletter that comes out every Friday afternoon.
