WATER WORK

by Marilyn Davis

In our bathrooms and laundry rooms, swimming pools and gardens, car washes and amusement parks, we depend on water, even revel in it--and, much of the time, take it for granted.

Not everywhere and not always, of course. Sometimes the tap threatens to run dry. During droughts, communities impose restrictions that put a crimp in our routines, reminding us that water is a finite resource, after all. But human activity, not just nature's, can run us into water trouble:

• The Ogallala aquifer, an underground reservoir that underlies much of the High Plains, is nearly exhausted in many areas due to decades of irrigation.
• So many water districts siphon water from the Colorado River that in some years, the river's delta at the Gulf of California is dry.
• The remaining extent of the Everglades is threatened by the diversion of water for agriculture and urban use. A massive project is now planned to restore some of the region's original hydrology.

Will the United States have enough water to sustain population and economic growth while protecting the environment? Policymakers can't get a good handle on that question without a solid understanding of today's water usage and a reliable way to predict future needs.

"It's a federal priority to create water-use science--a scientific way of estimating and analyzing water usage," says Ben Dziegielewski, professor of geography at SIUC. "We're trying to understand how water is used and what drives demand."

The business of water

Dziegielewski has studied urban water conservation and water-use forecasting for more than two decades. Last year he and SIUC economics professor Subhash Sharma finished a large study of water-use trends in the United States from 1950 to 1995. Funded by the U.S. Geological Survey (USGS) through the Illinois Water Resources Center at the University of Illinois, the study aimed to analyze historical water-use data, then use that information to create statistical models for predicting future water demand.

Since 1950, the USGS has compiled water-use data from the states every five years. This endeavor, called the National Water-Use Information Program, aims to include all surface-water and groundwater withdrawal points--a massive task, since there are more than 2 million nationwide, not counting many rural residential wells.

To make sense of all this data on water withdrawals, the SIUC team had to put it in the context of economic, geographic, and demographic information. Then they modeled the four "sectors" that account for most water use in the United States: irrigation, thermoelectric (power-plant cooling), public supply (essentially, urban use), and industry, in that order.

Industry sounds like it would be a big user, but it accounts for the smallest use among these sectors, due in part to declines in manufacturing over the past 20 to 30 years. Irrigation and thermoelectric account for four-fifths of the water withdrawals in the United States. They are followed by public-supply use, which surpassed industrial use in the 1980s.

Water use climbed 142 percent from 1950 to 1980. Over that time, the U.S. population increased only 77 percent. But 1980 was a watershed year, if you will--the peak year for water use in the nation. After that, water use leveled off and then began to decline. By 1995, it had dropped almost 10 percent from 1980 levels--even though our population grew 16 percent during that time and gross domestic product increased by more than half.

The team's findings show that power plants, irrigation, and industry all were using less water overall in 1995 than in 1980. Some of the drop was due to bad economic news: the decline of industries that are heavy water users, such as steel mills and other manufacturing concerns. But some was due to water conservation--in particular, more efficient irrigation techniques.

"We are learning to do more with less, and the USGS data bears that out," says Dziegielewski.

The Jacuzzi culture

The story is different for public use, however, where the demand for water has continued to increase. Domestic use nearly tripled from 1950 to 1995. Population growth was a big part of this jump. But so was higher per-capita use, resulting from a higher standard of living.

Simply put, people who have more money use more water. As people earn more, they can afford not just appliances like dishwashers, but also swimming pools, hot tubs, Jacuzzis. They put in nice landscaping that has to be watered. Expensive cars get washed more than beaters do. And so it adds up.

Water districts can't do anything about population growth, but they can about per-capita use. "In states where water is a problem, there are aggressive conservation programs to use water more efficiently," Dziegielewski says.

"If we pursue conservation in a dedicated way, it works," he adds. "Urban conservation programs that are carefully designed and properly funded are very successful in reducing water usage. Those that are not have mixed results."

Some conservation measures are voluntary; others, such as the low-flow toilets that are a perennial gripe of humor columnist Dave Barry, can be and have been mandated. But where we really lag behind, says Dziegielewski, is in using water pricing to discourage waste.

It's unlikely that we'll ever see full-value pricing, he says, but price "often isn't used in a market-oriented way." He adds that, ironically, water in the arid West is considerably cheaper, on average, than water in the East. Yet research, including his own, has shown that higher prices for water do work to decrease usage.

The states vary in how well they're doing to conserve water. Not all have seen declining water usage since 1980. Seventeen states have seen increases, despite the overall nationwide decline. Population growth, higher per-capita water use, and the need for more power plants were the key reasons. In Illinois, our public water-supply districts are drawing more water than ever.

A handful of states--among them, California, Texas, Illinois, New York, and Florida--exert a huge influence on national trends because of the sheer amount of water they use. Illinois ranks sixth in population, but third in water withdrawals.

What boosts our ranking? We're number one in water withdrawals for power-plant cooling--17 billion gallons a day. ( More water facts...)

A flood of data

Many of the factors affecting water use are obvious. The complexity lies in identifying them all and determining to what extent they contribute under varying conditions. Forecasters need to know which are most important to take into account.

"We wanted to see what the best predictors of water demand are," Dziegielewski says. The SIUC research team found that most historical variation in water use could be accounted for by changes in a few factors--what Dziegielewski calls "key explanatory variables."

For instance, they determined that 69 percent of the growth in per-capita domestic use of water from 1950 to 1995 could be attributed to a higher median family income. "Personal wealth is the major contributor," Dziegielewski says, "and the percent urban population is also a major contributor. The rural population just doesn't use as much water around a house."

Statistical modeling using multiple regression allowed Dziegielewski and Sharma to estimate the contributions of the variables they identified. They then fine-tuned the models by "backcasting"--testing them against known water use in 1980 to see how well they would have predicted demand.

To develop the models, Dziegielewski and Sharma needed historical data and projections on everything from weather to demographics to pumping costs for irrigation. Finding and analyzing that data required a small squadron of undergraduate student workers and graduate assistants, as well as the full-time work of geography staff researcher Thomas Bik (who is also a doctoral student). Besides the USGS data, they relied on data from the U.S. Census Bureau, Department of Labor, Department of Agriculture, Department of Commerce, National Climate Data Center, and other agencies, as well as nonprofit groups such as the American Water Works Association.

To grasp the complexity of the team's task, consider just one water-use sector, thermoelectric. To fully understand what governs water use by power plants, the team had to factor in data on type of plant (coal, nuclear, etc.), capacity, number and capacity of cooling towers at those different types of plants, annual air temperature, cooling degree-days and heating degree-days, details on state water law, and more.

By running statistical analyses, they were able to identify which factors were most important in explaining past water use by this sector. To forecast future use, they also had to factor in economic and demographic trends.

Sharma, a leading econometrician, did a lot of "diagnostic" work on the modeling to make sure the team's estimated values were unbiased--that is, free of statistical errors that would tend to cause under- or overprediction. Based on the backcasting to 1980 data, he and Dziegielewski were able to determine that, across all sectors, their models' overall margin of error was about 20 percent.

"That's not extremely accurate," Dziegielewski says, "but it's good enough to show that the data we collect nationwide can be used to identify variables we can change to be more efficient in water use. And water managers can use the models to refine predictions over time."

A look to the future

What do Dziegielewski and Sharma expect by the year 2040?

Irrigation use, they forecast, will drop by about 18 percent--a big help in conserving water resources. But that will be offset by increases in other areas. Industrial use will rise by about 20 percent and thermoelectric use by about 39 percent. Meanwhile, domestic use will soar by 61 percent.

Overall, the researchers expect U.S. water use to go up by about a fifth--21 percent--between 1995 and 2040, to 440 billion gallons per day. That's the same amount used nationally in the peak-use year of 1980.

Given a projected population of 377 million by 2040 (a middle-of-the-road estimate), the research team's forecast isn't such bad news. But many water districts will face tough times--including some that have not had supply problems in the past.

Dziegielewski is not an alarmist about water use in the United States, but he acknowledges that some parts of the country will see changes. For example, agriculture in parts of the High Plains, particularly southwestern Kansas and the Texas panhandle, will not be sustainable much longer at current levels. "The farmers know this, and many are already divesting," Dziegielewski says.

Illinois has a wealth of water resources--major rivers, lakes, and aquifers, along with decent rainfall. Yet we too need to pay attention to water conservation.

Droughts have triggered periodic water-conservation mandates in some Illinois communities. And the Northeastern Illinois Planning Commission predicts that a number of townships in the Chicago metro area will face water shortages by 2020.

There, the problem is that water withdrawals from Lake Michigan are capped, but urban sprawl continues. Communities in the region are increasingly turning to groundwater, including shallow aquifer systems that may not be able to meet demand sustainably. That is, the rate at which nature replenishes them may not equal the rate at which we deplete them.

Any water-use predictions must weigh many uncertainties about society, the economy, and nature itself. One of the greatest uncertainties is climate change.

Since the SIUC team was looking more at near-term predictions, Dziegielewski says, "we didn't include any scenarios for climate change. But we did include two variables--precipitation and average summer temperature--that really capture the effects of climate on water use. For example, we know that every 1 percent increase in temperature results in a .8 percent increase in water used [by communities].

"Given unfavorable climate outcomes--meaning less rain and more heat--we would expect up to 20 percent more water demanded for urban use."

Technological advances are another big unknown, but tend to balance out, Dziegielewski says: some increase water use, whereas others decrease it.

Standard of living and population are both projected to increase, and therefore to increase the demand for water. "The good news is that water use has been growing at a slower rate than population," Dziegielewski says. "The conservation ethic is also increasing. Perhaps we can offset demand with [voluntary] conservation measures, although that's hard to push forward."

Thomas Bik, the geography researcher who works with Dziegielewski, notes, "The predictions were based on the assumption that the future will look a lot like the present--or, here is what water use will look like if we keep doing what we're doing in terms of management. There's plenty of room for improving efficiencies in all sectors--especially if pricing starts to drive innovation."

The fact that Illinois is a leader in water-use science can help us in policymaking. Dziegielewski cites the expertise of the Illinois State Water Survey and the Illinois Water Resources Center, as well as a tradition of water resources research in SIUC's Geography Department stretching back some 30 years. (See An Alphabet Soup for H2O.)

"Water-use science is a young science, and we're trying to build it," he says.

Bringing it all back home

Dziegielewski is now applying what was learned from the USGS study to the Midwest, and Illinois in particular. In one project, he and Bik are developing water-demand predictions for all 102 Illinois counties for 2005 through 2025. They are working with Ken Hlinka of the state water survey, which is funding the research with a grant from the Illinois Department of Natural Resources.

"An increasing number of areas in Illinois, mostly urban areas, will be looking at potential difficulty in providing enough water for economic and population growth," Dziegielewski says. "We want to identify counties where demand is outpacing supply."

In 2004, he, Bik, and graduate student Xiaoying Yang will wrap up another study forecasting water needs over the same time period for public-supply use in six Midwestern states, including Illinois. That project is underwritten by the Midwest Technology Assistance Center for Small Public Water Systems with funds from the U.S. Environmental Protection Agency.

The work focuses on infrastructure: will water districts have not only enough available water, but enough treatment capacity to supply customers' needs down the line? "The idea is to help communities plan better for water supply," Dziegielewski says.

He and Jack Kiefer, a doctoral student in geography, also recently landed a grant from the Illinois Water Resources Center to take an in-depth look at water rate structures in the state and how they've changed over the past 20 years.

Few government agencies gather data on water rates, and no comprehensive study of water prices in Illinois has been done since the late 1960s. Yet most water system managers in the state rank rate-setting as their top concern. (That's according to a 1999 study by Dziegielewski, Bik, and SIUC agricultural economist Roger Beck.) Price, after all, is one of the few ways that water districts can influence demand for water.

Kiefer and Bik have surveyed some 1,600 water systems in Illinois to see what rate-making structures are in effect (and why), and what factors have driven rate changes. Kiefer is devising ways to measure "acceptability" criteria for rates. Among those criteria: Are rates fair and equitable for customers? Do they encourage conservation? Do they provide stable and sufficient revenue? Are they politically acceptable?

Kiefer will crunch the numbers and analyze them for his dissertation. Findings will be posted on the SIUC Geography Department's web site. Water utilities will be able to use them to gauge the benefits of alternative price structures. Consumers will be able to use them to see how customer preferences affect water rates--and whether rates are geared to ensuring an adequate supply of water.

"The basis of water management is to look for decision variables that are in our power to change, in order to conserve water," Dziegielewski says. "Price may be the most important.

"We can send a signal about the true cost of water without overcharging people. We shouldn't pretend that water is cheap."

Fall 2003 Contents | Perspectives Home | SIUC Home