Drinking Water & Health Newsletter
and Risk Perceptions
By Bruce Bernard
Drinking Water Blues
By Joan Rose
Rules Would Have Major Impact on Water Treatment Facilities
Are Disinfection By-Products?
Techniques For Minimizing Disinfection By-Product Formation
Risk and Risk Perceptions
by Bruce Bernard
How often have you been asked questions such as, "how safe is this
medication?" or how dangerous is skiing?," or, more importantly, "how safe
is our drinking water?" At the heart of these questions is a recurring
theme: risk. Risk has become an important concept for those concerned with
public health, particularly those charged with ensuring the delivery of
safe drinking water. The public today has a heightened sense of risk.
Those of us in the public health community, therefore, need to understand
the processes that contribute to public perceptions of risk, so we can
What is Risk?
The term "risk" may mean significantly different things to
different people. For this discussion, risk is defined as encompassing
both the likelihood of something negative occurring and the consequences
of that occurrence. For example, if thunderstorms are a common occurrence
in community X and thunderstorms in community X typically result in
significant damage, then the risk of damage during the next thunderstorm
will be high.
Risk assessment is a field of scientific research dedicated to
defining and measuring risk. Risk assessors calculate the likelihood of
particular events resulting in a specific consequence. The
event/consequence may be demonstrable or theoretical. If demonstrable, the
event/consequence has occurred before and one can calculate the likelihood
that it will happen again by reviewing documented evidence, such as, one
of every five thunderstorms in community X has resulted in significant
Calculating the likelihood of an occurrence based on a theoretical
assumption can be even more complex, though typically resulting in minute
probabilities. How likely is it that a terrorist will steal a nuclear
missile and point it at the United States? The fact that many Americans
may be afraid of this particular scenario illustrates that extremely
minute risk probabilities do not always translate into a public perception
of minimal risk. This situation also can be illustrated by examining
current debates over the safety of our drinking water. Recent concerns
over theoretical health risk posed by drinking water disinfection
by-products (DBPs) have raised fears about the safety of tap water. This
has occurred despite the fact that drinking water disinfection has for
years provided U.S. citizens with the safest drinking water in the world
protecting citizens from the demonstrated risk of waterborne disease.
At issue in the drinking water debate is an assessment of the
relative risk of two different circumstances the demonstrated risk of
waterborne disease and theoretical risk of disinfection by-products. The
process of making such an assessment is known as comparative risk
For years, regulators and risk assessors have evaluated risk based
on a risk-to-benefit analysis (i.e., does the benefit outweigh the risk).
However, in many instances the benefits are not quantifiable, are unknown,
or are based upon an individual's personal values. Since risk cannot be
eliminated but only replaced by other hopefully lower risks, a different
type of analysis is more realistic. Approximately 15 years ago, I coined
the term Comparative Risk for the process of comparing one risk to another
(i.e., which risk is lower given approximately equal or unquantifiable
benefits)? For example, when we cross the street, we quickly compare the
risk of jaywalking with the risk of losing our jobs if we are late for a
meeting. When evaluating a situation based upon comparative risk, we seek
to find a balance between two risks: If I'm late for this meeting I will
lose my job, so I will jaywalk. But I'm going to wait until that speeding
bus passes before I run across the street. Comparative risk analysis also
plays a major role in the drinking water debate.
Water treatment facilities across the country have a responsibility
to ensure the delivery of safe drinking water. Each day these facilities
seek and find a balance that minimizes both the risk of waterborne disease
and the risk of disinfection by-products. However, much of the public does
not understand this concept. Many argue that disinfectant levels should be
lowered to achieve better control of the remote risk of disinfection
by-products even though the scientific community warns such a move could
drastically increase the demonstrated risk of waterborne disease. Why does
the public nevertheless grow anxious? Because, as discussed above, remote
probabilities of risk do not always translate into a public perception of
Perceived Risk, Actual Risk and Acceptable Risk
As the drinking water debate illustrates, perceived risk is often
very different from actual risk. Risk anxiety levels are not directly
proportional to actual risk probabilities. Emotions and cultural values
often play a major role in distorting abstract scientific facts. And risk
assessors tell us that most individuals find self-imposed or
self-controlled risks (such as jaywalking) to be less risky. Familiar,
detectable and natural risks also are included in the category of
Conversely, risk assessors also tell us that circumstances
perceived as involuntary, new, undetectable, man-made or uncontrollable
are far less acceptable. Indeed, Paul Slovic, a physiologist from the
University of Oregon who studies public perception of risk, notes,
"Americans today feel they are at more risk from technology than ever
before. Yet, in terms of health, life expectancy, and even accidents,
things have improved greatly."
What confounds the dilemma of perceived risk versus actual risk is
the publics' sense of outrage when they feel that the people they have
trusted with their health and safety appear to be ignoring or minimizing
their concerns. It does not seem to matter whether perceptions of risk are
based on fact or emotion, citizens need to know that their health and
safety are being protected. Blind assurances that the risk of X happening
in community Y is so remote that there is no reason to worry are
Message to Our Readers
All of us in the public health community have an obligation to
understand and care about the concerns of our publics. Perceptions of risk
are not always based upon demonstrated evidence but rather emotion and
societal values. Whatever they are based on, the concerns of your
constituents must be addressed in a manner that reassures them that you
share their concerns and are continually looking for better ways to
safeguard their health.
Risk communication is an intriguing combination of science and
emotion. The following articles will provide more information on this
THE DRINKING WATER BLUES
By Joan Rose
College of Public Health University of South
After decades of taking it for granted, drinking water received
national interest when the media, the water industry and congressional
leaders turned their attention to one of the largest waterborne outbreaks
documented in the United States in the last decade.
Cryptosporidium, a protozoan parasite similar to Giardia,
infiltrated the city of Milwaukee, resulting in nine deaths and causing
between 211,000 and 400,000 cases of diarrhea. While Milwaukee wondered
when their tap water would be safe to drink again, public health officials
across the country were asking, will the public have confidence in the
drinking water? Is our drinking water supply at risk? Could this happen in
A relatively newly recognized pathogen (first identified in 1907),
Cryptosporidium has caused waterborne outbreaks in the United States
before: in Texas from a contaminated well (1985); in Georgia associated
with a filtered surface water supply (1987), and in Oregon from springs
and filtered river water (1992). Several waterborne Cryptosporidium
outbreaks have been documented in the United Kingdom as well.
Researchers have learned several things from these episodes.
Cryptosporidium is an intestinal protozoan and is transmitted through
fecal contamination. It may originate with animals (cattle, in particular,
are suspected). But the organism may also originate from human wastes and
has been found routinely in waste-water. The organism's environmental life
stage, the oocyst, can be commonly found in U.S. and Canadian surface
waters. The greater the level of waste input from humans and domestic
animals and the less watershed protection, the greater the concentrations
During an outbreak, some significant contamination event apparently
overwhelms the drinking water treatment system. In Milwaukee, domestic
waste-waters and animal waste discharge into the Milwaukee River, which,
in turn, discharges into Lake Michigan, very likely contributing to the
contamination. At the treatment facility, the situation was exacerbated by
less-than-optimal coagulation and filtration, as well as by recycling the
backwash waters from the filters. These circumstances, in addition to
changes in the removal efficiency of the filters, left the population of
Milwaukee vulnerable to Cryptosporidium, previously an unrecognized
Many utilities and communities throughout the United States and
Canada face a similar risk. What can be done? There are three important
and interdependent components in the process of producing safe water (1)
source protection, (2) filtration and (3) disinfection.
already have been initiated to evaluate potential sources of
Cryptosporidium in surface supplies which may be amenable to a
prevention control. The American Water Works Association has recommended
that water supplies be tested for the oocyst. This data will help
decision makers evaluate sources, and take into account the seasonal
water quality changes which may contribute to contamination events.
filtration is an effective step for the removal of the oocyst,
optimization and treatment reliability must be maintained.
- Finally, no single disinfectant will control this protozoan. In
the aftermath of the Milwaukee incident, ozone was mentioned as a
possible alternative to chlorine; however an outbreak of Cryptosporidium
recently occurred in Ontario, Canada, where ozone was used as a
pre-disinfectant. Additionally, if ozone is used, it must be used in
tandem with another disinfectant such as chlorine or chloramine because
ozone does not provide a residual disinfectant level. Together, this
dual disinfectant barrier could be very effective against most microbial
Despite having the safest tap water in the world, contaminants such
as Cryptosporidium have occurred in U.S. and Canadian drinking water
supplies. Those of us responsible for protecting public health and
providing safe drinking water must rise to meet this latest challenge: to
prevent future Cryptosporidium outbreaks and to assure the safety and
integrity of our nations' drinking water.
"REG-NEG" RULES WOULD HAVE MAJOR IMPACT
WATER TREATMENT FACILITIES
The Regulatory Negotiation (Reg-Neg) committee established by EPA
to negotiate and propose a new rule setting limits for drinking water
disinfectants and disinfection by-products has reached a tentative
agreement that may if implemented, encourage water treatment facilities to
move away from chlorination, the nation's primary drinking water
disinfectant for the past 80 years.
Safe Drinking Water Act
Under the Safe Drinking Water Act (SDWA), EPA is required to
establish new standards for drinking water contaminants; drinking water
disinfectants and disinfection by-products are considered contaminants for
the purpose of this rulemaking. Disinfection by-products sometimes form as
a result of a reaction between drinking water disinfectants and organic
matter (referred to as organic precursors) that occur naturally in many
However, some of the recent outcomes of the Reg-Neg process are
highly controversial. In fact, the National Rural Water Association, one
of five water industry groups invited to participate in the negotiations,
recently withdrew from the Reg-Neg process because of the lack of
science-based evidence in support of commit- tee recommendations.
The lack of supporting scientific evidence is most noticeable when
reviewing the details of the committee's proposed phase two
recommendations. Phase two calls for drastic reductions in maximum
contaminant levels (MCLs) which, if implemented, would require a major
overhaul in the nation's water-treatment infrastructure. If the message
being sent to the water industry is that a major overhaul is required,
documented evidence should clearly show that such an overhaul is both
necessary and beneficial.
The proposed disinfectant/ disinfection by-products rule would be
implemented in two phases. Phase one of this rule would lower MCLs for
chlorination by-products known as trihalomethanes (THMs) from 100 parts
per billion (ppb) to 80 ppb by 1997. At the same time, phase one would
establish a 60 ppb MCL for previously unregulated chlorination by-products
known as haloacetic acids. Phase one of the proposed rule also would
establish maximum residual disinfectant levels (MRDLs) for chlorine and
Additionally, the first phase of the rule would require large water
systems (those serving more than 100,000 people) over the next several
years to conduct raw (untreated) water monitoring for microbial
contaminants and a range of disinfection by-products in treated water. To
address unknown disinfection byproducts from all disinfectants, organic
carbon precursor removal will be required if total organic carbon (TOC) of
the finished water is greater than 2.0 mg/l.
If implemented, certain phase one requirements would apply to both
large and small water systems. Phase one compliance costs are estimated to
be in the range of $3 billion in capital, and over $400 million in
annually recurring operating and maintenance expenses.
The proposed disinfectant/ disinfection by-products rule (effective
in the year 2002) specifies further lowering of MCLs to 40 ppb for total
THMs (TTHMs) and 30 ppb for haloacetic acids. The stringent MCLs
established in phase two, referred to as a "backstop," will automatically
become law unless the second regulatory negotiation (REGNEG 2), which will
occur in 1998-99, reaches agreement on some other levels. The Chlorine
Institute, which has been represented on the RegNeg Technologies Work
Group, has objected to the phase two recommendations.
Clearly, both phases of the disinfectant/disinfection byproduct
rule proposed by the Reg-Neg committee would have a significant impact on
water utilities throughout the United States. The new rule would be
expensive and may require the adoption of alternative water treatment
technologies. Many of these alternative technologies may have adverse
public health and environmental impacts of their own. For example,
ozonation is seen by some as a viable alternative to chlorination.
However, in water sources containing bromide, ozone forms byproducts known
as bromates, which scientists recognize as a far more significant health
risk than THMs or haloacetic acids. In addition, ozone breaks down quickly
and does not provide disinfection through the water system to the tap.
It is vital that all concerned with the nation's public health take
steps to ensure that the final disinfectant/disinfection by-products rule
be based on sound science and reflective of logical and attainable public
All interested parties should be sure to comment on the proposed
rule, which is expected later this year. For more information on the
RegNeg process, contact the U.S. EPA Safe Drinking Water Hotline at
WHAT ARE DISINFECTION
Drinking water disinfection has been, and continues to be, one of
the most important public health measures of the 20th century, having
greatly reduced the threat of waterborne disease in the United States,
Canada and other developed countries. Over the past 20 years researchers
have been looking at the by-products of reactions between drinking water
disinfectants and organic matter that exists naturally in many water
sources. Resulting compounds are commonly referred to as disinfection
Increased media coverage of disinfection by-products, paired with a
heightened sense of risk amongst the general public, has caused more
people to ask about the safety of the nation's drinking water. (See
related story on Perceptions of Risk). Most often, they look to the
nation's public health community for answers.
While the body of science regarding these issues is inconclusive,
those responsible for the nation's public health and the safety of its
drinking water must be able to accurately respond to constituents
questions about these issues. The following overview should help.
How Are Disinfection By-Products Formed?
When raw water (from a lake, stream or other untreated source)
enters the water treatment system, it typically carries organic materials
it picked up along the way, such as decaying vegetation and animal wastes.
Some of these materials serve as organic precursors and react with
disinfectants during the treatment process, forming disinfection
by-products. These by-products occur more frequently in the summer, when
the level of organic material in raw water typically is higher, and water
temperatures generally increase.
Disinfectants also can react with other naturally occurring
constituents in raw water, such as the bromide ion, to form bromine-based
or brominated disinfection by-products (see chart below).
Disinfection By-Products - Formation
Assessing Potential Human Health
halocetic acids, aldehydes
Chlorate ions, chloride
Bromide (from sea
awater, oil field brine or geological weathering of minerral
Does not oxidize
Does not oxidize
compounds (Iron, manganese)
Does not oxidize
minimal quantities of chlorate
Other Factors that Affect By-Product Formation
1Disinfection by-products occur from a reaction between
disinfectants and certain source water characteristics. This chart
identifies by-products from different disinfectants and source water
Several other source-water characteristics affect the quantity and
type of disinfection by-products that may form during the disinfection
process. They include:
Assessing Potential Human Health Risk
Disinfection by-products were first identified in the early 1970s.
Much of this research has focused on chlorination by-products since
chlorine is used in some form by over 95 percent of water systems in the
United States. This research has been enhanced by the development of
sophisticated laboratory techniques that can measure chlorine compounds in
water in parts-per-billion and even parts-per-trillion. Likewise, these
techniques have raised questions about other disinfectants, which also
create byproducts. However, the body of knowledge about by- products from
alternative disinfectants, such as ozone, is incomplete.
Trihalomethanes, or THMs, are a class of chlorinated organic
compounds that sometimes form as a result of the chlorination process.
Early studies suggest that these compounds are carcinogenic to certain
laboratory animals at levels much higher than commonly found in drinking
water. Many studies have been undertaken to determine whether THMs
constitute a human health risk.
In 1990, the International Agency for Research on Cancer (IARC), an
arm of the World Health Organization, assessed the strength and
reliability of virtually every major scientific analysis of the potential
health effects of chlorinated drinking water. They found that chlorinated
drinking water was not a classifiable human carcinogen.
Moreover, various procedures exist to remove organic materials from
water during the treatment process, thereby reducing disinfectant
byproduct formation (see related story).
TREATMENT TECHNIQUES FOR MINIMIZING
Treatment techniques are available that provide water suppliers the
opportunity to maximize drinking water safety and quality while minimizing
the risk of disinfection by-product formation. Disinfection by-products
typically form as a result of a reaction between naturally occurring
organic matter in raw or untreated water and drinking water disinfectants.
One of the best methods to control disinfection by-products is to
remove the organic precursors prior to disinfection. Other conventional
methods include changing the point of chlorination and lowering chlorine
feed rate. An October 1991 American Water Works Association (AWWA) Water
Quality report identified effective procedures for reducing the formation
of disinfection by-products, as follows:
Organic Precursor Removal
There are three ways to effectively remove organic precursors:
Coagulation and Clarification
Most treatment plants use
the coagulation process to remove as much sediment as possible.
Coagulation processes can, however, be used effectively for natural
organic matter removal. Precursor removal is possible when aluminum or
ferric salts are used as coagulants for sediment control. Further
precursor removal is achieved by reducing the pH prior to or during the
addition of these coagulants.
A recent study by Reckhow and Singer (Reckhow, David A., and Philip
Singer, Chlorination By-Products in Drinking Waters: From Formation
Potentials to Finished Water Concentrations) notes that alum coagulation
"has long been known as an effective means of removing natural organic
matter." The study also states that alum coagulation is effective for
removing organic precursors. Numerous studies show that ferric salts also
are very effective.
Adsorption processes have been used
successfully in some applications for removing disinfection by-product
precursor material. Activated carbon historically has been used to provide
adsorption both granular activated carbon and powder activated carbon are
available for this function.
Historically, membranes have been
used for desalination of brackish waters. One process, which uses pressure
to force the liquid through a membrane, has proven particularly successful
in removing disinfection by-product precursors. The AWWA report states
that membrane procedures "actually remove precursors from the finished
product, which makes it a promising alternative for future control of
Conventional Treatment Optimized for Disinfection By-Product
There are two ways to optimize conventional treatment:
Change the Point of Chlorination
According to the AWWA
Water Quality Committee report, "one of the simplest solutions for the
reduction of THM formation is the movement of the first point of chlorine
application to as late in the treatment process as possible." By moving
the point of chlorination to late in the process, most organic precursor
materials will be removed before the water is chlorinated, thus minimizing
disinfection by-product formation potential.
Lower the Chlorine Feed Rate
By lowering the rate at
which chlorine is applied to water, there is less tendency for the organic
precursors to react with chlorine, thus reducing the potential for
World Health Organization Reports "No End in Sight" for Global
An additional 131, 000 cases of cholera including 2,265 deaths,
have been reported so far in 1993 and the epidemic shows no signs of
abating, the World Health Organization (WHO) recently reported.
More than half of these new cases were reported in Peru and
approximately 44,000 were reported in four African countries Malawi,
Mozambique, Zambia and Zimbabwe.
A new cholera bacteria has been found on the Indian subcontinent
and is likely to spread, according to WHO.
The ongoing epidemic has ravaged Latin America and Africa since
1991, resulting in over 600,000 cases of cholera and 7,000 deaths.
Drinking Water & Health Newsletter is a Publication of
the Public Health Advisory Board to the Chlorine Chemistry
Safe Water Advisory Committee
Sanford M. Brown, Jr.
School of Health and Social
California State University, Fresno
Bruce K. Bernard, PH.D.
National Consumer League
Galveston County (Texas)
Pan American Health