JUST AS WE NEED SIGNIFICANT volumes of water for energy, we also consume a lot of energy for water. We use energy to store, pump, treat, move, clean, and use water. Much of the water we use is for irrigation or power plant cooling. Usually the water for irrigation doesn’t need to be treated beyond simple filtering, if at all. Power plants will treat water to prevent minerals from accumulating inside equipment, but they do not need the water to meet potable standards. Drinking water for our municipal systems typically requires extensive treatment. Then, after we use the water in our homes or businesses and flush it down the drain, it has to be treated again to raise it to a standard that is safe to return to the environment without causing damage to the ecosystem. The full supply chain includes pumping and conveyance from the original water source, water treatment, water distribution, end-use, wastewater collection, wastewater treatment, then discharge. Generally speaking, the treatment steps are very energy intensive, although there is great variability. Despite their requirements, heating at the end use is the biggest energy consumer of all.1
Though abundant globally, water is often out of reach. When water is far away from where we live or is dirty, we spend a lot of energy moving it, cleaning, or storing it. The idea that energy is needed for pumping water is certainly not new. The ancient Egyptians widely used Archimedes’ screw, a clever device that uses the manual turning of a screw to elevate water. The tight coil of connected blades could keep raising water as long as it was continually operated. Its invention is attributed to Archimedes in the third century BCE, but it might have been in use even earlier. It is still employed today, in modern water-lifting stations at amusement parks, water treatment plants, and elsewhere. It turns out that robust designs are still useful thousands of years later.
The reason that water needs so much energy for pumping is because it is so dense: it weighs 8.34 pounds per gallon. That density is also one of the reasons why water is so valuable as a coolant and a process material. That also means it takes a lot of energy—and therefore money—to bring that water uphill. Marc Reisner’s classic book Cadillac Desert captures this idea from the people living in deserts who begrudgingly noted that water moves uphill toward money and power.2
The energy needed for pumping water depends on how far the water needs to be raised, the rate at which it is raised, pipe diameter, friction, and so forth. The energy that is needed to raise water up out of a well is the energy that is required to overcome the force of gravity, which wants to pull the water back down. Raising a 2.75 gallon bucket of water (about 10 liters) a distance of about 330 feet (100 meters) requires approximately ten Btu (or 10,000 Newton-meters) of energy. Pumping the water up only 33 feet (10 meters), for example by raising the water from a river to the top of a nearby riverbank, requires one Btu (or 1,000 Newton-meters) of energy. That doesn’t seem like much, but for large volumes, the energy adds up, and the distance it has to be elevated drives the energy needs.
For continuous flows, it is more useful to look at the power that is needed. Pumping 2.5 million gallons per day (29 gallons per second, or 110 liters per second)—enough water for 1,900 average Americans—out of an aquifer 330 feet below the ground requires 107 kilowatts of pumping power. Keep in mind that a typical house needs 1–3 kilowatts of power on average to run the whole place, so a pump that size consumes the same power as approximately thirty to one hundred homes.
A medium-sized U.S. city with a million residents might need about 150 million gallons of water per day. Raising that water from surface sources to elevated water treatment plants over a height of one hundred meters requires a little more than 6 megawatts of pumping power. Massive wind turbines are approximately 1 megawatt apiece, so that city would need a half-dozen of them running full-bore just for pumping water up from its source to the top of the hill, after which it can flow downhill to customers. That means pumping water uphill requires a lot of electrical power, which requires a lot of money. And money is a proxy for political power. So Reisner’s book was correct: water moves uphill toward power and money.
It takes a lot of energy to lift water. The Archimedes screw (top) is one mechanism for doing so. Energy is invested to rotate the screw, which raises the water, as at Schlitterbahn, a water park in New Braunfels, Texas. [Photo by Jeffrey M. Phillips]
Imagine pumping that water by hand, the way it once was. This energy intensity is one of the reasons why people like shallow wells and surface water.
And all those energy and power requirements for pumping water to higher elevations don’t include the energy for treating the water to make it potable or pumping the water through pipes to our homes. For cities that have water treatment plants at high elevations, much of the distribution can be done by gravity, but not all cities have that advantage. Cities whose layout does not accommodate gravity-fed distribution with natural geography instead pump the water into ubiquitous elevated tanks that look like giant golf balls sitting on a tee.
In addition to the energy to convey the original source of water to the treatment plants, the treatment processes themselves require a lot of energy. Water and wastewater treatment require that energy for a variety of actions, including pumping, blowing, aeration, ultraviolet lamps, stirring, and the embedded energy in the chemicals that are added to the process. Words like flocculation, filtration, carbonation, and sedimentation are standard parts of the water treatment engineer’s lexicon.
The amount of energy needed to treat water and wastewater to a suitable form depends on a variety of factors, such as how contaminated the source water is, the nature of the contamination, what the water will be used for, plus the physical features and treatment approach of the facility. Dirtier water generally requires more energy for treatment, and end-uses that require high standards of cleanliness also need more energy. Hospitals, semiconductor cleanrooms, and food preparation facilities need water that is much cleaner than what is required for cooling industrial equipment or irrigating farms. That means there is great variability nationally for the energy intensity of water treatment.
Bottled water is perhaps the most energy intensive use of all.3 Energy is required to process, bottle, seal, and refrigerate the water, but it turns out that the energy that goes into making the plastic bottle itself is the biggest piece. And, if that bottle is moved by trucks, planes, or ships over long distances, the energy consumption goes even higher. In modern cities that have well-functioning piped water systems that draw from local sources, the water is produced in a comparatively energy-efficient and affordable way. In those situations, the extra money and energy for bottled water seems like a wasteful indulgence that offers little additional value. However, after a natural disaster such as a hurricane wipes out a local water system, or in developing countries where water systems are contaminated or do not exist at all, bottled water can be a lifesaver.
Considering all the energy used for water treatment, the national average is 1,400 kilowatt-hours per million gallons for sourcing, treating, and distributing surface water.4 The water used for electricity is often denoted in gallons per kilowatt-hours, while the energy for water can be tracked in kilowatt-hours per million gallons. Energy for conveyance varies from zero for gravity-fed systems, as in New York City, to approximately 14,000 kilowatt-hours per million gallons for Sierra snowmelt water delivered via the State Water Project from Northern California over mountain ranges to Southern California destinations such as San Diego.
It turns out that the State Water Project, California’s ambitious water conveyance project that moves water by aqueducts across the region, singlehandedly consumes 2–3 percent of the state’s electricity.5 The project would make the Romans proud. The world’s largest water pumps were invented just to lift the water two thousand feet over the Tehachapi Mountains. That shipment of water is still a source of contention within California today. The total length of the State Water Project is about seven hundred miles, serving 25 million people and consuming 5 billion kilowatt-hours each year. The Central Valley Project has about five hundred miles of canals, consumes 1 billion kilowatt-hours annually, and includes eleven hydroelectric power plants to harness electricity from the snowmelt. A lot of that water goes to Southern California cities plus agricultural operations growing nuts and fruits. During the peak of the California drought in 2015, it was a widely noted complaint that it takes a gallon of water to grow an almond and several gallons to grow a walnut in the Central Valley of California, which provides nearly the entire nation’s supply of domestic nuts.
When the water is sourced from groundwater, the national average is higher, at 1,800 kilowatt-hours per million gallons for sourcing, treating, and distributing it, primarily because of the additional energy for pumping the water to the surface. Brackish groundwater requires even more energy, with a range of 3,900–9,750 kilowatt-hours per million gallons. That range depends on the level of total dissolved solids in the water: the more salts that have to be removed, the more energy is required. At the high end of the scale is seawater desalination, with a range of 9,780–16,500 kilowatt-hours per million gallons. Desalination is very energy intensive, and its varying energy requirements are a function of water temperature and salinity. Although desalination systems are robust and drought-resistant because the world is awash in saline water, they are very energy intensive. In addition, they produce brine streams of water that are even saltier and require disposal. Both the energy consumption and the disposal represent environmental impacts that are important to consider.
There are basically two standard approaches to desalination that are widely implemented: thermal systems and membrane systems. Those two approaches provide about 95 percent of all desalination globally, with other minor approaches such as freezing and electrodeionization providing the rest.6 Thermal systems include actions as simple as boiling water, which separates out the salts by evaporating the water, which can subsequently be condensed. This is just distillation as learned in high school chemistry. This system is very effective but requires a lot of energy. Thermal approaches also include more sophisticated systems like multistage flash, which uses a series of steps and pressure drops to evaporate water, and multieffect distillation. Thermal systems can be very crude, as they can use any source of heat. This crudeness means they have low efficiency in practice, but it also means they can use waste heat as the source. So when thermal desalination systems are integrated into large facilities with a lot of waste heat, such as power plants, then the overall efficiency is improved. Membrane systems use a semipermeable barrier to filter out the dissolved solids. The most popular membrane approach is reverse osmosis. Other membrane approaches include forward osmosis, nanofiltration, and electrodialysis.
Many desalination facilities use reverse osmosis, pushing salty water through cylindrical membranes that separate the salt from water. Doing so creates two outputs: a stream of less salty water (permeate, or product water) and a stream of saltier water (concentrate, brine, or reject water).
Reverse osmosis filters use electrically driven pumps to push water through membranes that separate out the salts. The membranes separate two streams of water with different levels of salinity. Normally, osmotic pressures would drive salts from the side with greater salinity toward the side with lower salinity as a way to achieve equilibrium, so that concentrations on both sides are about the same. The pumps reverse this direction, causing the salty side to get saltier and the fresh side to get less salty. Reverse osmosis systems are generally more energy efficient than boiling water, but are expensive technologies that require electricity as opposed to just heat.
After treating water and getting it to our homes, we use it to flush away our wastes. In a book aptly titled The Big Necessity: The Unmentionable World of Human Waste and Why It Matters, the author Rose George humorously describes her adventures going around the world and writing about how different people relieve themselves.7 She examined the practice in the United States, United Kingdom, Japan, China, and India, revealing some interesting insights along the way. For something so basic to human nature as emptying one’s bowels, there is a surprisingly varied range of approaches. Even the phrase as simple as “going to the bathroom” is uniquely American, as we do not necessarily have baths in most of the rooms where we use the toilet. In the United Kingdom and Europe, the bathroom actually means the room with a bath in it, so instead they use the word “toilet” or “water closet” or “loo.” For the United States, the “toilet” is the actual commode, whereas in Europe, “toilet” refers to the room that holds the commode. The French have the phrase “toilet water” to denote perfume. The Japanese have the world’s most luxurious toilets, with robotic functions, heating, built-in cleaners, and all sorts of options. Many rural Chinese have outhouses or latrines. And in India, many people defecate openly in the fields.
In addition to observing the differences in the act of defecation, Rose George also follows where the waste goes. She travels down into the sewers and to the wastewater treatment plants to see the technology and the investments required by society to make them work. The understated conclusion from all of her writing is that waste matters. Waste is a differentiator of society. Societies that deal with their waste are healthy and wealthy. Societies that do not manage their waste are sick and poor. The cause and effect is not clear: whether societies are healthy and wealthy because they manage their waste, or whether they manage their waste because they are healthy and wealthy is not obvious, but the correlation is hard to miss.
Managing wastewater and sewage is a big task, and it is all part of what we term “sanitation.” The system is pulled together by sewers or drainage pipes that bring the solid-laden flows to the wastewater treatment plant. Sewers themselves can be a massive infrastructure undertaking, but then the wastewater treatment plant itself is where the real hard work begins. As my colleague Ashlynn Stillwell likes to say, “When we flush waste ‘away,’ it’s not gone. Wastewater treatment plants are the ‘away.’ It’s that place no one likes to think about. It’s where all our waste ends up.” When we took our students on a tour of a wastewater treatment plant, we told them we were taking them away. In the animated movie Flushed Away, the setting is the historic sewers of London, further reinforcing this notion that away and sewers or wastewater are synonymous. Given our distaste for our own waste, “away” is an apt description for where we want it.
Wastewater treatment plants are usually at a low elevation so that the sewage can flow by gravity to the facility. Consequently, most of the energy requirements are used for the treatment process as opposed to conveyance. On average in the United States, wastewater treatment requires 955–1,900 kilowatt-hours per million gallons, depending on how many stages of treatment are applied and the level of cleanliness.8 However, based on local prevailing standards of treatment, the requirements can be higher. In Austin, Texas, the wastewater treatment plants require 2,700 kilowatt-hours per million gallons. One reason is that Austin is the only city in Texas, and possibly one of only a handful in the United States, where the river segment downstream of a major metropolitan area has a better water quality designation than the segments upstream. Raj Bhattarai, one of the principal engineers at Austin Water Utility, beams with pride when he gets a chance to tout this surprising fact. As a native of Nepal high in the Himalayas, he understands the importance of protecting watersheds for the benefit of people downstream.
What this all means is that the water system overall requires a lot of energy. And for municipalities, it is usually the single largest energy tab for the city government because water and wastewater treatment facilities are usually municipally owned. It is typical for a water and wastewater utility to consume 50 percent of all the electricity used by a city’s government. The energy intensity of water and wastewater treatment remains an opportunity for conservation.
One key conclusion from all of these energy requirements is that treating water, treating the wastewater that is produced, then treating the effluent from the wastewater treatment plant again to potable standards is less energy intensive than desalination. Closing the loop and using treated effluent is one option for reducing the energy requirements for the water system. This process is called “toilet to tap,” and it grosses out a lot of people. While the idea of toilet to tap may be distasteful, it is a fine way to provide water.
That approach is already in action indirectly when we discharge our effluent to a river or aquifer for the next city downstream to drink. Andy Sansom, a noted water expert, is fond of saying, “We are all downstream from somebody.” Houston is drinking the wastewater from Dallas. Belgrade is doing the same with wastewater from Vienna. But somehow we feel better when the water goes through “nature” first, as opposed to staying in pipes the whole time. While nature does provide some cleaning services in many instances, it is not obvious that nature’s cleaning is better than what our engineered facilities can achieve.
“Toilet to tap” works. Singapore built such a facility in 2000, calling it NewWater. The system provides 60 million gallons per day or 30 percent of the drinking water from reclaimed wastewater.9 NewWater has worked fine and is slated to triple its capacity by 2060. Water recycling and toilet-to-tap treatment systems are also valuable for the military. Shipping water to the front lines is an expensive and deadly proposition. The long supply convoys span thousands of miles and are soft targets for enemies, meaning that the cost of water per gallon ultimately is several orders of magnitude higher as delivered to the military theater than at the local grocery store or from our taps. Consequently, the U.S. military spends a great deal of money testing and deploying on-site water treatment systems that can make potable water from degraded streams. These systems cost a lot of money and require extra space up front to be shipped to the forward operating base, but after that, they spare the need for a lot of shipments of water, saving lives.
The International Space Station also has a reclaimed water system to produce drinking water because the cost of shipping water to space is exorbitantly expensive.10 It costs from $10,000 to $90,000 per pound to ship cargo into space. That means shipping freshwater to astronauts in space costs anywhere from just under $100,000 to nearly $750,000 per gallon. Because of the high costs for freshwater, the space station collects and treats the graywater from washing, urine, and condensed moisture from breath and sweat to be drunk again. The station does not have any blackwater, as the astronauts do not use toilets.
Interestingly enough, my research at Stanford University for my Ph.D. was a part of this project to create an onboard water treatment system for the space station. As a graduate researcher in mechanical engineering I invented and deployed laser-based sensors that could measure a variety of trace gases. One of my patents from that work was for a sensor that would measure very small concentrations of ammonia in the presence of other species, such as water vapor or carbon dioxide.11 I also studied combustion, the cornerstone of our modern energy system, giving me an early insight into energy conversions, consumption, and impact from the perspective of a thermoscientist. I used the sensors I invented to measure emissions of pollutants in the flue gases from combustion systems. In particular, I looked at the unwanted emissions of ammonia that slip out of the smokestack after the ammonia had been injected to scrub pollutants out of the stack.
Smokestacks have a lot of carbon dioxide and water vapor. The ability to measure trace quantities of ammonia in the presence of carbon dioxide and water was a useful performance advantage. Just like smokestacks, humans breathe out a mixture of carbon dioxide and water. That means the ability to measure ammonia in a smokestack is good preparation for measuring ammonia on the space station in the air mixed from the exhalations of astronauts.
At the time, NASA was developing an on-board water treatment system that would turn the graywater into drinking water. I flew to NASA’s Johnson Space Center with my ammonia sensor in hand to make some measurements of their space-bound water treatment system that was undergoing ground testing. Bringing the whole energy-water nexus story full circle, Johnson Space Center was named after Lyndon Johnson in honor of his enthusiastic support for the space agency while he was vice president in the early 1960s. This is the same President Johnson who pushed for the Rural Electrification Act that helped poor families far from cities pump water into their homes, much as the “Farm Woman’s Dream” poster envisioned.
It was through this experiment that I teamed up with professors, postdocs, and students at Rice University. Dr. Frank Tittel was the principal investigator of the project and the person who extended the invitation to me for the collaboration. Through that project I interacted briefly with Dr. Bob Curl, who won the Nobel Prize with Rick Smalley at Rice University. This is the same Rick Smalley whose top ten list of grand challenges facing humanity included energy and water at the top. Unfortunately, I never met Smalley.
NASA had set up a test station with a large bioreactor that would convert streams of wastewater laden with organic and nitrogenous components into fresh, potable water. Making freshwater available onboard for the astronauts is a key life support mission and had remained a vexing challenge for decades, so this experiment was considered an important priority. We were tucked into a spacious, nondescript experimental facility on the campus with large pieces of equipment next to a bathroom. NASA employees, including scientists and support staff, were invited to use the urinal, sink, and shower to provide the sample wastewater materials the bioreactor would convert into drinking water. It was not unusual for commuters to ride their bikes to work in the morning and then hop into the experimental shower to freshen up before they started working. Then there was a steady stream of volunteers throughout the day who came by to use the urinal or toilet.
The blackwater from the toilet was sent to the conventional wastewater system, but the graywater was sent to the experimental water treatment system that we were evaluating. Because there was no blackwater to manage, the treatment requirements were less intensive. However, the graywater still had a lot of organic matter from the soap, skin cells, and dirt, and nitrogeneous matter from the urea in the urine. Furthermore, since sickness in space is hard to manage, the water standards still had to be stringent to avoid propagating any waterborne illnesses.
The water treatment system was a simple configuration of a glass tube vertically aligned with ceramic shells inside, which the graywater would flow through. The shells would accumulate some of the undesired contaminants and would force mixing with the treatment chemicals. The treatment process would produce off-gassing of water vapor, carbon dioxide, and ammonia. The carbon dioxide was from the organic contaminants and the ammonia was produced from the urea and other nitrogenous compounds. By measuring the fraction of ammonia mixed in with the water vapor and carbon dioxide, I could help the scientists assess the relative health of the bioreactor.12
A schematic of the NASA Advance Water Recovery System for water processing aboard manned spacecraft. I invented an ammonia sensor for my doctoral dissertation that was used to measure off-gassing from the biological waste processor.
I spent two weeks at NASA in August 2000 integrating my sensor into the experiments and taking data. With that experiment complete, I defended my dissertation less than three months later. One decade after I conducted those experiments, the water treatment system was finally launched in 2010.
While the energy needs for pumping and treatment are significant, it turns out the energy needs for preparing water at the end of the pipe are even larger. These energy investments are for processes such as filtering, chilling, pressurization, deionization, and heating. In particular, water heating is very energy intensive. It shouldn’t be a surprise. Water requires so much energy for heating that it is used as the standard for energy units. In English units, one Btu (British thermal unit) is the energy required to raise the temperature for one pound of water by one degree Fahrenheit. In standard international units, one joule (J) is the energy required to raise the temperature for one gram of water by 0.24 degrees Celsius.
And water heating is a useful application for energy. After all, hot water serves many helpful purposes, including disinfection and sterilization at hospitals and washing dishes at home. Not to mention that hot showers (and therefore hot water) are a lot more comfortable than cold showers. So water heating is an important proxy for quality of life.
Including all of these end uses drives up the total energy consumed for water dramatically.13 By adding up all these different uses of energy for water, appliance by appliance and sector by sector, a total national estimate for energy consumption for water can be deduced. Dr. Kelly Twomey Sanders, a professor in civil engineering at the University of Southern California, did just that. In fact, she was the first person to tackle that research problem on the national scale with so much accuracy. She found that the United States consumed approximately 12.3 quadrillion Btu of energy just for water in 2010. That year, the United States consumed 98.0 quads of primary energy, which makes water directly responsible for about 13 percent of consumption. Another 34 quads was consumed to generate steam for indirect purposes, such as process heating, space heating, and electricity generation. That last one is particularly impressive: we spend a remarkable amount of energy each year just boiling water at power plants to make steam that spins turbines to make electricity.
Of the 12.3 quadrillion Btu we consumed as a nation for water services and direct steam use, a little more than a fifth was just for heating water in our homes and businesses. On-site water pumping was relatively low in the residential sector in comparison to the industrial and commercial sectors, as housing units tend to be smaller. Residential water systems for most single-family residences operate off the prevailing pressure of the water distribution network, so pumps are seldom needed at all. But, for tall residential buildings, water must be elevated to tanks at the top of the building after which it is fed by gravity to the individual units. Those residential buildings need on-site pumping to raise the water. Large industrial facilities also require large quantities of energy to move water around on-site.
The total amount of energy we spend on water is impressive. It means as a nation the United States spends more energy on water than for lighting. And therefore the energy embedded in water is big enough to care about. The Natural Resources Defense Council and Pacific Institute referred to this phenomenon in their landmark 2004 report for California as “Energy Down the Drain.” Literally, the energy embedded in cleaning, pumping, and heating water gets lost down our drains.14 It also means that water conservation is a reasonable pathway to energy conservation. But interestingly, water conservation has been mostly ignored, perhaps because water is so cheap. By contrast, we have spent a lot of political energy as a nation fighting about lightbulb standards. President George W. Bush signed into law the Energy Independence and Security Act of 2007, which mandated an efficiency standard for lightbulbs. The standard basically prohibits the continued sale of conventional incandescent lightbulbs, whereas compact fluorescent lightbulbs and light-emitting diodes meet the requirements. Years later the standards remained a contentious issue and were removed by congressional Republicans as part of budget negotiations to keep the government open in December 2014. Similar fights have been waged over fuel economy standards for automobiles. The whole time, water has been ignored as an option for energy savings, despite its large energy footprint and significant opportunity to enable energy conservation.
