When the well’s dry, we know the worth of water. —Poor Richard’s Almanack, 1746
ONE OF THE MAIN CONSEQUENCES of the interdependence of the two systems is that water constraints can become energy constraints and energy constraints become water constraints. Because the power sector is so water intensive, it is particularly vulnerable to water constraints. In fact, water can be too hot, too cold, too abundant, or too scarce for full operation of power plants, leaving a sweet spot where the water needs to be just right. That means heat waves, freezes, droughts, and floods all cause problems for power plants.
Heat waves (disregarding the drought that often accompanies heat waves) hamper the power sector in two primary ways: they reduce performance because of reduced power plant efficiency, and they put power plants at risk of violating thermal pollution limits. As an unavoidable consequence of the Second Law of Thermodynamics, power plants and automobile engines are less efficient when it is hot outside. That means the hotter temperatures of the heat wave degrade the efficiency of power generation. The efficiency of a power plant drops a percentage point just for the case where the water is ten degrees Celsius hotter from a heat wave. That might not sound like much, but for a nuclear power plant with one gigawatt capacity, a 1 percent loss of efficiency can cost more than five thousand dollars per day in lost revenues, because there is less electricity to sell.
In addition to the performance problems, environmental regulations have an effect as well. Because power plant cooling structures in the United States are regulated by thermal pollution standards from the Clean Water Act, there is a limit on the maximum exit temperature allowable for the return water from a power plant’s cooling system. Typically those requirements, which are designed to protect the aquatic environment, are fixed at a particular temperature, say 104 or 112 degrees Fahrenheit. In some cases, the maximum allowable return temperature for the cooling water is based on the difference between the inlet and outlet. When a heat wave occurs, the temperature of the incoming cooling water will increase above normal. This puts the power plant at risk of returning water to its source at a level that exceeds its thermal pollution threshold, so it must intentionally cut back its operations to avoid a violation.
Extremes in water temperature and availability do not work well with power plants. The power sector needs water to be in a sweet spot—not too hot, too cold, too scarce, or too abundant—to function at full power.
Extreme heat waves that affect the power sector can happen. The 2003 heat wave in France was one prominent example. That heat wave stretched across Europe, but France bore the brunt of its impact. In many places, the temperature anomalies in France were ten degrees Celsius (eighteen degrees Fahrenheit) hotter in 2003 than in 2001.1 That was a killer heat wave, showing some of the health risks from climate change. According to one scientific paper, the heat wave killed approximately 70,000 people in Europe.2 France in particular suffered nearly 15,000 to 20,000 additional deaths. In particular, the elderly were vulnerable. The demand for power was spiking as people turned on their air conditioners to avoid dying from the heat.
At the same time that the demand for power was peaking, the French nuclear fleet had to draw down its output to avoid violating the thermal pollution limits.3 Because the French power sector has a very high contribution from nuclear power with water-intensive open-loop cooling systems and because the French nuclear plants are often sited on inland rivers (thirty-seven of their fifty-eight nuclear power plants are so situated), they are doubly exposed to heat waves. By contrast, ocean water maintains a much more stable and cooler temperature compared with inland rivers and lakes.
Because of the combination of the higher energy demand and higher river temperatures, as many as seventeen of France’s fifty-eight nuclear reactors were at risk of violating the thermal pollution limits and had to reduce their output or turn off completely.4 The rivers were too hot and the water levels too low to guarantee adequate cooling of nuclear power plants, putting the entire system at risk of failure. Just as demand for electricity was peaking for air-conditioning in response to the heat and with life hanging in the balance, power supplies were being cut back because of the very same heat wave. Ultimately, Electricité de France (EdF), the main power provider in France, requested exemption from its operational limits and cut its power exports in half to keep the power system operating. After those temporary exemptions were granted, nuclear facilities were allowed to operate but at reduced capacity. In total, the nuclear power fleet had to dial back its output by up to 15 percent for five weeks. France and many other regulatory bodies in Europe “overrode their environmental laws and allowed for higher waste water temperatures.” When making the choice between impinging the environment through overheating of the ecosystem and saving human lives, decision-makers selected the latter.
Though the heat wave in France was certainly not a normal situation, the fear with climate change is that those killer heat waves will become the new normal. And, the phenomenon is not restricted to France: because of unseasonably warm weather conditions in May 2012, the Illinois Environmental Protection Agency (IEPA) granted Exelon’s Quad Cities nuclear power plant a provisional variance from its temperature limits.5 As in France, the thermal pollution limits were temporarily voided during a heat wave to keep the power on.
Ocean water is generally cooler and its temperature more steady than inland rivers, but even seaside power plants are at risk of shutting down from heat waves. A few months after the Quad Cities plant needed a variance, the Millstone Power Station, a nuclear power plant in Connecticut on Long Island Sound, shut down for two weeks because of overly hot seawater, which exceed the 75-degree Fahrenheit temperature limits at the inlet of the cooling water system.6 As before, the authorities did their best to shunt aside environmental regulations to keep the power on, but despite an emergency license amendment from the U.S. Nuclear Regulatory Commission to show a lower temperature by averaging different measurements, the safety threshold was violated, triggering a full shutdown.
Even though water can be too hot—a problem that might be more prevalent with global climate change—water can also be too cold. It is known by power plant operators that weatherization and preparing for cold storms includes steps such as draining nonessential water systems to avoid damage from ice formation, and confirming that the water in essential water systems was actually circulating, which reduces the likelihood of a freeze. So in effect, power plant operators know that freezing water poses a risk to reliability.
In an illustrative incident, Texas endured a statewide freeze in February 2011. That cold snap caused water to freeze in some instrumentation pipes, tripping some large coal plants off-line, which triggered a cascading series of power plant failures elsewhere, ultimately leading to rolling blackouts across the state.7 On one day, six more coal-fired units at four locations went off-line. Then another old coal-fired plant went down when “an old switch filled with water and froze, causing it to break,” adding to the strain.
Because the natural gas demand was at a record high to meet the heating requirements for homes and business, backup power plants fueled by natural gas did not have enough gas available in the pipelines. Ultimately, more than fifty gas plants were not able to work. Compounding things, many gas-producing wells use electrical equipment, and some gas compressors along pipelines use electrical pumps. When the power went out, those wells and compressors stopped working, making it even harder to meet demand and keep the pipelines filled with gas, causing pressures to drop and straining the system further.8 Even if the pumps had not turned off because of the power outage, some gas wells turned off anyway because of a phenomenon known as “freeze-offs,” which occurs when the water produced alongside natural gas freezes or crystallizes, blocking off the gas flow, effectively shutting-in the well.
Ultimately, between February 1 and 4, 2011, more than two hundred individual generating units within the Texas grid experienced an outage, were derated, or failed to start.9 All of these events combined triggered statewide rolling blackouts: since enough power could not be produced to meet demand, demand had to be cut across the state. The grid operator, sitting in a control room, starting turning off the power to different parts of the state one after the other. Some parts lost power for only twenty minutes, whereas others lost power for up to eight hours.10 Over 3 million customers—and there are usually several people per “customer,” which is synonymous with “electricity meter”—lost their power, and a total of 4,000 megawatts of capacity was shed. The after-action report that investigated the blackout episodes concluded that a vast preponderance of the failures were from frozen water in sensing lines, equipment, and water lines.11 Who knew that little old pipes frozen with water could bring the whole grid crashing down?
In the wave of polar vortices that occurred in early 2014, there were a few additional instances of frozen water disturbing the energy system. Above Niagara Falls, concerns about ice damaging the hydroelectric power plant made headlines, causing authorities to bring out ice-breaking ships to protect the equipment and to allow the water to spin the turbines again.12 Ice also prevented thermal power plants from working. During the same cold wave, ice on Lake Superior prevented coal boats from delivering fuel to power plants. The U.S. Coast Guard had heavy-duty icebreakers working overtime trying to clear lake ice, and ice cutters escorted the coal ships. Because of blocked deliveries, the Presque Isle power plant was running out of coal and bought 25,000 tons of excess stockpiled coal from a nearby power plant operated by Marquette Board of Light and Power.13 Because there were no trains or barges between the two locations, the coal had to be transported by truck, which is an unorthodox and expensive way to move coal.
And, in one series of cascading failures that same winter, frozen water in the Midwest undermined the energy system, creating life-or-death situations.14 Rivers and lakes froze, which prohibited the passage of barges carrying road salt. Because road salt couldn’t get delivered, roads couldn’t be deiced. Because the roads were icy, diesel trucks carrying propane had trouble delivering heating fuel to customers freezing to death in the upper Midwest. Making matters worse, earlier in the year, a relatively wet growing season yielded a bumper crop of corn for biofuels that was wetter than normal. Consequently, propane demand was very high as the fuel is used to dry the corn for long-term storage. In this case, excess water caused higher energy demand for drying corn, after which frozen water made it harder to deliver salt, which meant frozen roads inhibited deliveries of propane. So a water problem (icy rivers, too much rain for the crops) became an energy problem (propane shortages) exacerbated by another water problem (icy roads), which became an even bigger energy problem (disrupted energy delivery). The energy delivery system is more fragile than people think.
In addition to water temperature affecting the energy sector, the abundance of water also matters. In particular, water scarcity from extended drought can be a problem. Though hydroelectric power is attractive for many reasons, it is least reliable during droughts when the need to use water for other purposes such as drinking and irrigation might take precedence over hydroelectricity. For example, low water levels in hydroelectric reservoirs can force power plants to dial back or turn off.
Utilities reduce hydroelectricity generation when water scarcity occurs, and those reduced flows can also lead to higher rates for electricity. In some places, the risks are quite acute.15 Although in the United States this phenomenon is an inconvenience, in other parts of the world, it’s a tradeoff between power and quenching our thirst. For example, cities in Uruguay were confronted with the choice during the 2008 drought of whether they wanted the water in their reservoir to be used for drinking or electricity. Ultimately, reduced hydroelectric power output can strain the entire grid in regions where hydro is a primary power source.16 The Energy Information Administration’s monthly update in April 2012 raised an alarm that the California grid would face a number of challenges in the summer, partly because of the lower-than-normal snowpack in the Sierra Nevada mountain range, which was expected to reduce hydroelectric output.
Returning to the example introduced earlier, in July 2012, the grid did come crashing down—in India—because of strains induced by decreased hydroelectric output.17 This event was the largest blackout in history, affecting more than 600 million people. Drought, triggered by the later-than-normal monsoon season, played a role in two ways. First, there was reduced hydroelectric power, making it harder for the grid to meet demand. That problem was exacerbated by floods months earlier that caused runoff from farms to silt up behind dams, reducing their capacity.
Second, because of reduced rainfall, the need for electricity from farmers to pump water for irrigation was higher than normal, straining the grid. There were simply more farmers using pumps to move water, which drove up demand, causing different regions of India to overdraw compared with their allotted power. Because of a long history of overpumping combined with reduced recharge to aquifers, water tables have been falling for years in India. Farmers using groundwater for irrigation had to lift water over a higher elevation gain, which also increased energy demand. Furthermore, farmers using surface water had to pump the water greater distances to cover a greater swath of land, also increasing energy demand. All of those responses combined to cause spiking demand for power at the same time that hydroelectric generation was lower, pushing the system past its breaking point. Once the power was out, millions of pumps for wells used by farmers and households quit working, as did the water supply and wastewater treatment systems for municipalities. Thus, a water event (drought) caused an energy event (increased demand) that caused another energy event (power outage) that caused a water event (water supply shortages). Ultimately, the energy constraint became a water constraint.
The problem of water scarcity is not just limited to hydroelectric reservoirs: it also affects thermal plants that need vast amounts of cooling water. This problem was particularly acute during the extensive drought in 2007 and 2008 in the southeastern United States. Unlike the problems of heat waves and thermal pollution, which might cause a power plant to reduce its power output, if water levels drop below the physical location of water intake pipes, then the power plant ceases operating completely. During the drought in the southeastern United States, lake levels came within three and a half feet of the water intake pipe for the cooling system at the Harris reactor near Raleigh, North Carolina, and within one foot of the intake level needed for one of the backup systems at the McGuire plant near Charlotte. It is difficult to expand, lower, or lengthen the water intake pipes so that they are less exposed to falling lake levels. According to one AP report,
Extending or lowering the intake pipes is not as simple at it sounds and wouldn’t necessarily solve the problem. The pipes are usually made of concrete, can be up to 18 feet in diameter and can extend up to a mile. Modifications to the pipes and pump systems, and their required backups, can cost millions and take several months. If the changes are extensive, they require a [Nuclear Regulatory Commission] review that itself can take months or longer. Even if a quick extension were possible, the pipes can only go so low. It they are put too close to the bottom of a drought-shrunken lake or river, they can suck up sediment, fish and other debris that could clog the system.18
Power plants are amazing testaments to human ingenuity and innovation. They are also hard, slow, and expensive to adapt because they depend on large physical structures.
Lake Lanier, a key reservoir in northern Georgia that provides drinking, irrigation, and cooling water, reached its record low level in December 2007.19 Consequently, power plants near Atlanta during the winter 2008 drought were within days of shutting off because the vast amounts of cooling water were at risk from diversion for other priorities such as municipal use for drinking water. At the same time, amazingly, there was no restriction on municipal lawn watering, which drained supplies from Lake Lanier even further. Overall, 24 of the United States’ 104 nuclear reactors at the time were sited in the region that endured that drought.
When drought and heat waves are combined, as they often are, the problems are compounded. For the same heat wave in France in 2003 described earlier, there was a simultaneous drought that caused France to lose 20 percent of its hydropower capacity.20 France gets 78 percent of its electricity from nuclear power and approximately 11 percent from hydropower, which made this drawback hard to accommodate. The same problems happen in the United States. In one instance in August 2007, the Tennessee Valley Authority shut down one of three reactors at its Browns Ferry nuclear plant because drought lowered water levels in the Tennessee River, which was its source of coolant. Low water levels with the hottest temperatures in several decades plus water warmed by upstream power plants heated the available water above acceptable thresholds. The same river’s tributaries were also low, reducing the output from neighboring hydropower plants.
The drought lowered the water levels in the rivers, which made them more vulnerable to increased temperatures from the heat wave, causing a violation of the thermal pollution limits. The power problems were exacerbated by the fact that the same drought that caused problems for the thermal plants also reduced output from the hydroelectric power plants, causing upward pressure on power prices.
Drought causes other problems for the energy sector. Severe drought in 2012 threatened a two-hundred-mile stretch of the Mississippi River from St. Louis to Cairo, Illinois, with the risk of barge traffic being shut off in early 2013.21 The Mississippi is one of the world’s most important inland navigation routes, moving several billion dollars of goods each month. But because of the drought in 2012, the Army Corps of Engineers decided to save water for summer irrigation in its upstream reservoirs. Consequently, releases were reduced, diminishing flow to the Missouri River, which subsequently reduced the flow into the Mississippi River. If water levels get too low, barges heavily laden with goods would not be able to travel safely. Those barges carry things like fertilizers, salt (for winter road treatments—the same kind of salt that couldn’t be delivered over frozen rivers in 2014, causing energy shortages in the upper Midwest), and agricultural products. They also carry a lot of coal to midwestern power plants. That means a water scarcity event can threaten the supply chain for the energy industry in yet another way. And, in this case, water for irrigation was deemed more critical than water for the power industry. It is hard to imagine shipping the tonnage of coal and refined fuels by truck instead of barge, just because of a limit to the availability of trucks and capacity of the road systems. Furthermore, doing so would surely drive up prices dramatically.
In addition to water being too hot, too cold, and too scarce, it can also be too abundant. It is hard to imagine that water abundance can be a problem, given the prior discussion about the risks that droughts pose to energy production. But too much water also puts power plants at risk of damage. In one case a Nebraska nuclear power plant nearly shut down because of flooding of the Missouri River in June 2011.22 The floodwaters came within a few feet of cresting over temporary, inflatable flood walls surrounding the facility. One of the berms collapsed, triggering concerns of the need for a shutdown to prevent safety risks. And, for the Fukushima incident in Japan in 2011, it was not the earthquake that caused the biggest problems. Rather, the tsunami and the subsequent flooding ruined backup safety systems and triggered the meltdowns and explosions within the nuclear power plants. Those same floods can also overwhelm water and wastewater treatment systems, causing a spike in electricity demand.
The problems that water constraints cause for energy are not restricted to just the power sector: they also affect fuels production. The problems of water show up in a variety of ways. For extracting minerals such as coal, water constraints are not as prominent. Water use for washing and dust control does occur for coal, but it is relatively minor in terms of the overall production, and that means coal production is not threatened very severely by water shortages. Conventional oil and gas production use water-flooding in some reservoirs to enhance productivity, and so in those cases, water scarcity might restrict production.
Biofuels, because they require so much water for their growth, are the most sensitive to water constraints. Floods also destroy energy crops and cause massive loss of soil and runoff. Droughts hurt agricultural production in general, and energy crops such as corn and soy for ethanol and biodiesel are not spared. When the epic drought of 2012 hit the corn belt, corn production dropped and prices for ethanol increased. Also, some cities, including even Champaign, Illinois, in the middle of the corn belt, feel that they are in competition with agricultural users of water for energy crops.23
New techniques like hydraulic fracturing are also vulnerable to water constraints. Many of the sites with booming shale production, such as the Eagle Ford Shale in south Texas and the Bakken Shale in North Dakota, are in areas that are drought-prone. The shale producers may find themselves angering locals as a result. Because the oil and gas producers are often the newest local users of water compared to the entrenched users such as agriculture and municipalities, they arouse a lot of resentment when they suddenly show up for an oil and gas boom and purchase a lot of water to meet their production needs. Some towns are already banning the use of municipal water for fracking. Officials for the Ogallala Aquifer included hydraulic fracturing when they approved restrictions on water use in the groundwater district.24
At the same time, the shale producers are willing to pay a much higher price for the water than conventional users. They can afford to pay higher prices because the oil and gas that are produced are so valuable, and by contrast the price of water is very low. According to the CEO of Breitling Oil and Gas, Chris Faulkner, the company paid $68,000 for 3.5 million gallons of water, which is 0.2 percent of the $3.5 million spent to hydraulically fracture the well.25 In total, water procurement, injection, collection, trucking, and disposal for fracturing operations can be as high as 10 percent of the total cost of an active drill pad.26 It is safe to conclude that the availability of water, not the price, would be a key constraint.
While the risk to fuel production is usually whether there is enough water available at the front end to perform the extraction, the more important risk for shale production might have to do with the wastewater that is generated at the back end of production.27 Shale gas wells produce significantly less wastewater per unit of gas that is ultimately recovered than conventional gas wells, but the abundance of wells in a small geographic area that has not historically had a lot of production means that the flows of wastewater can exceed the local capacity to deal with them. How effectively the wastewater from shale wells is managed will be the ultimate constraint on production.
The total wastewater from an unconventional well comprises three different flows. There are the fracturing fluids (or “frac fluids”) that are injected into the well. These typically include freshwater along with a mix of chemicals designed to enhance productivity through a variety of means, for example by reducing friction, chemical precipitation, scaling, and biological fouling and changing viscosity. These chemicals raise public concern because of the potential risks they pose to ecosystems and human health. Of the millions of gallons of frac fluids that are injected into the wells, a major fraction comes back to the surface as “flowback” water.
In addition, significant volumes of “drilling fluids” or “drilling muds” are brought back to the surface. Drilling fluids are used to cool and lubricate the process of cutting through the rock while also clearing out the cuttings by bringing them to the surface. These drilling fluids have high concentrations of total dissolved and suspended solids, which complicates their disposal. At the same time, because horizontal drilling expands the length of wells compared to conventional drilling, it is reasonable to expect a lot more drilling wastewater.
Last, water is naturally present in the shale formations. When gas is extracted from the wells, this “produced” water comes with it.28 Because the water is highly saline, with a density of total dissolved solids that can exceed 100,000 milligrams per liter or more, meaning it can be three times more salty than the ocean, it is sometimes referred to simply as “brine.” But, it has many more components than just salt, including high concentrations of other materials that are of concern, such as metals, organics, arsenic, and sometimes naturally occurring radioactive materials (called NORM for short, though most people wouldn’t consider them normal components of drinking water). While the “flowback” and “drilling fluids” are generated as wastewater during the drilling and fracturing phase of the well’s life cycle, the “brine” continues to be produced over the well’s lifetime.
In the Marcellus Shale, a typical well generated approximately 1.25 million gallons of wastewater, just over half of which was brine, one third was flowback, and the rest was drilling fluids.29 The unconventional wells produce about ten times more wastewater than conventional gas wells, but they also produce about thirty times as much gas. So while the total wastewater volumes are higher for unconventional wells, which can create local environmental hotspots, the wastewater-to-gas ratios are actually more favorable. The unconventional wells have a gas-to-wastewater ratio about three times higher than the conventional wells over the first four years of production.
These are still high volumes of wastewater, and when multiplied by thousands of wells, the question becomes how to handle the waste. As the number of wells increase, so does the total wastewater flow. Initially, the wastewater is stored in surface ponds, which pose a risk to water quality if the ponds are not properly lined because the dirty wastewater can trickle down to the freshwater aquifer belowground. After that, the management options are limited.
There are a few ways to handle the wastewater: treatment, reuse for subsequent production, and disposal. Because of the high solids content of the wastewater, conventional municipal wastewater treatment facilities are not well equipped to handle the streams. Consequently, the dirty wastewater can pass through without being cleaned up sufficiently before it is discharged to the local rivers, which can elevate their pollution levels beyond acceptable thresholds.30 Highly specialized industrial wastewater treatment facilities, which are more capable but also more expensive and energy intensive, are more appropriate. Unfortunately, the industrial facilities are rarely proximate to the drilling site, which means millions of gallons of wastewater must be transported by truck or pipe from the well to the facility. This movement of water is essentially an unintended interbasin transfer of water from one watershed to another. While that is not a big deal at small volumes, for many thousands or tens of thousands of wells, the volumes of interbasin transfer become nontrivial.
There is also the possibility for on-site treatment to reduce wastewater volumes. While this approach is an advancing technology, it is in limited use at this time. That approach, however, remains a critical technical solution to the problem, allowing producers to avoid trucking or piping their wastewater all over the state.
As discussed earlier, underground injection of wastewater is another method of disposal. While that method is common in Texas, it is uncommon in Pennsylvania (where it is essentially forbidden because of a lack of suitable injection sites). The disposal options for the Barnett and Eagle Ford Shales in Texas are therefore different than for the Marcellus Shale in Pennsylvania. Consequently, some of the wastewater from Pennsylvania was trucked over the state line to Ohio, where it is allowed to be injected. Unfortunately, deep injection disposal triggered a series of more than one hundred earthquakes in the span of just over a year, culminating in one on New Year’s Eve 2011 in Youngstown, Ohio, that was 3.9 in magnitude on the Richter scale.31 These temblors rattled Ohioans in a place where earthquakes are uncommon and building codes are not designed to meet seismic standards. What was particularly galling for many people in Ohio is that it felt as if Pennsylvania were reaping all the economic benefits of shale gas production while Ohio felt the brunt of its drawbacks.
It is also possible to dispose of the wastewater through “road spreading,” which is basically just what it sounds like. Admittedly, that does not sound like a very appealing option and is prohibited for Marcellus wastewater, although it does have some useful purposes such as ice and dust control. Other reuses include recycling the wastewater to use in subsequent wells. In fact, that approach, while more expensive today than injection, is growing rapidly and might be an effective solution to the treatment and disposal constraints. This method is not without its problems, however: the wastewater has high solids concentrations, which can inhibit production from shale gas wells that are second or further down in line. The presence of bacteria in reused water can trigger biological phenomena such as fouling and the creation of corrosive materials, which also downgrade production. In places where water for frac fluids and wastewater disposal are expensive, recycling is likely to be an economic solution.
The key challenge is that the sudden growth in wastewater volumes in production areas can exceed local infrastructure and capacity to handle it. So, the shale gas wells, popping up by the thousands, overwhelm the local systems, ultimately constraining production on the back end of the life cycle. In the final analysis, it has been determined that the risks from fracking are more acute for water quality than water quantity.32
Just as water limitations can restrict energy, energy limitations can also restrict water. One of the key risks to the water system is its energy dependence: because the water sector requires so much energy, energy disruptions can cause water disruptions, especially for the water and wastewater treatment plants. This phenomenon is particularly true for power outages. Electricity is not the only form of energy consumed for water—natural gas and petroleum are also used for water heating, for example—but electricity is the form of energy that is used for water pumping. That means when the power goes out due to storms or intentional acts, water treatment plants cannot pump water to serve their distribution networks and wastewater facilities quit functioning. For example, Hurricane Ike knocked the power out along broad swaths of the Texas and Louisiana coast, causing sewage pipes to stop flowing, creating a public health risk.33 The consequences can trigger severe public health impacts if tainted water makes its way through the distribution system or if sewage goes untreated and is discharged into waterways.
Usually water systems pump water into elevated storage tanks. That means after a power outage water can still be served to many customers by gravity. But those water tanks do not get refilled when the electric pumps quit working, meaning the water pipes will go empty eventually. While it is possible for the natural gas grid to fail—for example when demand is really high for residential heating—it is more likely that the electrical grid will be disrupted. The primary difference is that the gas grid is below ground and the electrical grid is mostly aboveground, making it vulnerable to high winds that knock the lines down or knock trees down into the lines.
The outages described earlier from Hurricane Katrina, which affected water supplies in New Orleans, make up another prominent example. The impact of power outages does not just have to be on a large scale: smallscale power outages can also be highly disruptive. For example, for houses that get their water from wells that operate with electric pumps, power outages become water outages.
I happen to live in a neighborhood in a suburban area of Austin that until very recently used well water to serve ninety-five homes, despite the fact that we are surrounded on all sides by conventional piped-water systems. In May 2011, a little less than a year before the small well with its electric pump was connected to the neighboring piped systems, it suffered a power outage. The well operators taped a notice to the front door at my house that warned us to boil our water prior to consumption “as a precautionary measure.” Even in modern times in rich cities in urban, industrial areas, we have to periodically boil our water to purge it of health risks such as water-borne pathogens, all because of an energy constraint. In this case, bringing the nexus around full circle: we use energy to fix problems with water that were caused by energy outages.
In the end, boiling water is pretty easy for someone like me, as all I have to do is turn the knob on my stovetop. If I do not want to do that, I can drive to the nearby grocery store to get some bottled water to meet my needs until the water service is restored. However, in developing countries, boiling water is a time-intensive, laborious task. While the energy outage is a mere inconvenience for me, it is a women’s rights or human rights issue in other parts of the world.
Though centralized power outages can knock out the water treatment system, distributed power outages in just a neighborhood or at an individual home don’t stop the entire water supply. Many people might have experienced a power outage at their homes during a windstorm. Even though the power goes out, in many cases the phone lines still work. In fact, historically people have needed to phone the power company to let them know the power was out. And, in almost all cases, the water stays on because it is fed by gravity into the home.
Adjusting to a power outage is inconvenient, but it can usually be managed with candles providing light and batteries providing backup power for computers and other electronics. It is annoying, but does not cause severe discomfort. The bigger inconvenience by comparison is a water outage. The patience to go without flushing toilets, operating sinks, or taking showers is much harder to come by than the ability to go without electricity.
The energy outages also create public health consequences. For example, after Hurricane Sandy created large surges of stormwater and knocked out the power at wastewater treatment plants, hundreds of millions of gallons of stormwater mixed with untreated sewage passed through and spilled into waterways.34 At least six treatment plants in the New York area alone shut down completely, and many others were impacted in some way or another. Because of the stormwater floods combined with the failed pumping systems, the wastewater treatment plants got backed up, causing the sewage to flow the wrong direction back out of the drainage pipes. In a scene reminiscent of the science fiction movie The Blob, where a mysterious biological blob attacks whole cities, the New York Times noted that in one neighborhood “a plume of feces and wastewater burst through the street like a geyser.” Raw sewage was still seeping into homes at least five weeks after the hurricane struck. As the media coverage noted at the time, that consequence might be one of the longest lasting legacies of the hurricane. For these treatment plants, the electrical equipment was flooded, causing them to switch off. Making these treatment plants more resilient will require elevating the electrical equipment above the flood line and waterproofing the equipment. Only in this way can cities avoid a water event that triggered an energy event that triggered a water event.