Some Global Environmental Issues of the 21St Century

Some Global Environmental Issues of the 21st century

RICHARD WILSON

Department of Physics, Harvard University

Cambridge, MA 02138, USA

Abstract

There are two complimentary approaches to studying the environmental effects of fuel use and energy conversion. The first is to be completely systematic and to follow through each technology in great detail, and itemize each impact, whether large or small. The second, followed here, is to pick out those items that either dominate the risk or impact, or dominate the public perception thereof, and try to understand them in some detail. I pick the following:

(a) What is the effect upon health of particulate air pollution at today's levels? Experts increasingly believe that particulates kill 70,000 people a year in the USA, but it has not yet been officially admitted by any government.

(b) I will discuss another global environmental effect of a heavy metal: the pollution of water supplies, particularly in the Bengal Basin (West Bengal and Bangladesh) by arsenic and the implications this has for the world.

(c) I will then argue that the future of the world demands administrative control of a number of environmental issues. This is most clear when we discuss high intensity radioactive sources and plutonium.

1. The issues of low dose linearity

I start with an introduction to the important issue of whether or not there is linearity of effect with dose at low doses. Since 1928 this has often been assumed to be the case for radiation, in order to ensure adequate radiation protection. I argue that it is probably true both for radiation and true for many other polluting agents. There is no doubt that high doses of radiation (500 Rems) have led to acute radiation sickness and death; and doses just less than these (100 Rems) have led to cancer. But there is much more doubt whether the low doses that arise from normal operation of nuclear power plants, and even doses to most of the people exposed in accident conditions, lead to any health problems. However, it is conventional, for prudent public policy, to assume that there is a linear relationship between dose and response so that even small doses, if widely spread over a population, can produce an appreciable response. This assumption, was originally suggested by Crowther [10] and for many years was only made by those concerned about radiation exposure. This led to an (incorrect) feeling that anything involving radiation is uniquely dangerous. It is now realized that this low dose linearity assumption is probably equally valid (or invalid) for other carcinogens, and even other medical outcomes. This is an inherent consequence of the multistage theory of carcinogenesis, particularly in the form developed by Doll and Armitage [2, 3]. Indeed the idea is more general: if medical outcome is indistinguishable from one caused by natural processes and the agent acts as the same way as the natural processes at any stage in the carcinogenic process, then almost any biological dose-response relationship will be differentially linear [11]. Recently it has been realized that linearity may apply to many other situations such as particulate air pollution also [9]. This is crucial as we reevaluate data on air pollution. In 1925, for example, the cry was ”the solution to pollution is dilution”, thereby bringing all concentrations below an assumed threshold. This certainly reduced local pollution, but increased pollution at a distance and made a local problem into a global problem - albeit one of smaller individual concern.

2. The Effects of Air Pollution

The effects of fossil fuel use on public health are primarily those of air pollution: the liberation of gasses from the power plant as a result of fossil fuel burning. Burning of fossil fuels results in emission of gases from incomplete combustion as well as gases from impurities. There is a marked difference between fossil fuel and nuclear plants in these respects. The emissions from fossil fuel plants occur in ordinary operation and are continuous, whereas the only important emissions from nuclear plants occur in accident situations. The problems of pollution of coal were noticed in the 15th century in England. In the 17th century Evelyn wrote a tract on the subject [13]. There is no doubt that air pollution and in particular the burning of coal HAS killed members of the public outside the power plant when air pollution levels were high. After a large fraction of people got sick and died in bad fogs in Meuse Valley, Belgium, and in Donora, Pennsylvania, people paid attention. Immediately after a London fog in December 1953 there were 4500 ”excess deaths” [5], (Figure 1). Deaths were due to a variety of causes all of which can also occur naturally as shown in Table 1. The British government took the immediate action of banning the burning of soft coal in the cities. The plentiful supply of oil from the Middle East enabled the UK economy to do without coal burning.

At that time the government’s scientific advisors, believing in a threshold, argued that (i) if pollutant levels could be reduced 5 fold the effect would vanish, (ii) the pollution only affected the aged and sick who would die in a few days or weeks anyway and (iii) the major problem was sulfur dioxide from burning of the sulfur pollution in the coal. It is my contention that all three scientific statements (i) (ii) and (iii) were wrong and there has been, and still is, a major controversy about (ii) in particular, and how to extrapolate these known hazards to the lower levels of today. During the 1970s, I and many others thought that present day air pollution in eastern US or northern Europe affects 1% of the people exposed [25]. But in an influential 1979 review [15] several prominent British scientists systematically discounted the evidence presented in the studies carried out so far, which merely compared average death rates with averaged outdoor pollution and are classified by epidemiologists as "ecological" studies. They concluded that the health effects of particulate pollution at low concentrations could not be "disentangled" from health effects of other factors. The 1982 EPA report which expounded the reasons for the change to a PM10 standard stated that "the relationship between long-term exposure to air pollution has been extensively studied, but few of the studies provide sound or consistent findings sufficient to make quantitative conclusions." A major (unstated) reason for dismissal of the findings was the (correct) observation that by 1980 visible air pollution in many major cities (London, Glasgow, Pittsburgh, Moscow) had already been much reduced over the black periods of the first half of the century. No government took any further action.

TABLE 1. Numbers of deaths assigned to various causes, London Administrative County:
Weeks ended November 29 to December 27, 1952*

* From a Report of the Ministry of Health.

† The fog covered the period Dec. 5-8, and the week ending Dec. 13 was the week of heaviest mortality.

‡ Excluding deaths at ages under 4 weeks.

However there is now much stronger scientific justification. Exposure of animals to combustion products begins to elucidate mechanisms [1]. Systematic correlation studies of death rates with air pollution in major cities show consistent results (Figure 2). Moreover two epidemiological cohort studies avoid many of the criticisms that applied to the ecological studies. The first was the "Six Cities study" [12] involving a 14-16 year prospective follow-up of 8111 adults living in 6 U.S. cities: Harriman, Tennessee, St. Louis, Missouri, Steubenville, Ohio, Portage, Wisconsin, and Topeka, selected to be representative of the range of particulate air pollution. Measurements were made of total suspended particulates (TSP), PM10, PM2.5, SO4, H+, SO2, NO2, and O3 levels. Although TSP concentrations dropped over the study periods, fine particulate and sulfate pollution concentrations were relatively constant. The most polluted city was Steubenville; the least polluted cities were Topeka and Portage. Differences in the probability of survival among the cities were statistically significant (P= 0.001). Individual health outcomes were compared with average exterior concentrations. Mortality risks were most strongly associated with cigarette smoking, but after controlling for individual differences in age, sex, cigarette smoking, body mass index, education, and occupational exposure, differences in relative mortality risks across the six cities were strongly associated with air pollution levels in those cities.

These associations, shown in Figure 3, are stronger for respirable particles and sulfates, as measured by PM10, PM2.5, and SO4, than for TSP, SO2, acidity (H+), or ozone. The association between mortality risk and fine particulate air pollution was consistent and nearly linear, with no apparent "no effects" threshold level. The adjusted total mortality-rate ratio for the most polluted of the cities compared with the least polluted was 1.26 [95% confidence interval (CI) 1.08-1.47]. Fine particulate pollution was associated with cardiopulmonary mortality and lung cancer mortality but not with the mortality due to other causes analyzed as a group. The results are substantially larger than those in one of the best earlier "ecological" studies. This difference suggests that the cohort study is able to estimate the effects of air pollution with more accuracy. Ecological (population) studies are averaging the observed effects over the affected population, therefore lower rates in these studies compared to cohort studies would be expected.

Similar results were observed in a second, larger prospective cohort study [24]. Approximately 500,000 adults drawn from the American Cancer Society (ACS) Cancer Prevention Study II (CPS-II) who lived in 151 different U.S. metropolitan areas were followed prospectively from 1982 through 1989. Individual risk-factor data and 8 year vital status data were collected. Ambient concentrations of sulfates and fine particles, which are relatively consistent indoors and out were used as indices of exposure to combustion source ambient particulate air pollution. Both fine and sulfate particles are used as indices of combustion source particulate pollution, which is considered by many to be a likely agent. Sulfate and fine particulate air pollution were associated with a difference of approximately 15 to 17 percent between total mortality risks in the most polluted cities and those in the least polluted cities.

Adjusted mortality-rate ratios (and 95 percent confidence intervals) of total mortality for the most polluted areas compared with the least polluted equaled 1.15 (1.09-1.22) and 1.17 (1.09-1.26) when using sulfate and fine particulate measures, respectively. A simple application of these results to the continental USA suggests that 70,000 persons die early (have their lives shortened) by air pollution. I assume that one third, or 20,000 deaths arise from the existing coal fired electricity generation (about 200 GW-yr) to get a coefficient of 100 deaths per GW-yr.

These and other, data are summarized in [26]. This, if true, dwarfs all other health problems of fuel use.

A general model which might describe the effect is that the air pollution reduces lung function in an irreversible way. Lung function falls with age, and in the presence of the pollution could fall to a dangerous level when all sorts of ailments occur, at an earlier age than otherwise. This is shown diagrammatically in Figure 4. It is easy to see geometrically, that the calculated ”loss of life expectancy” is directly proportional to the assumed lung damage, and death rate is also proportional. Figure 5 suggests that reduction in lung function is, on average, related to air pollution variables. This is a delayed effect that is not easy to affect after the initial lung damage, similar to the delayed effect of the cancer mortality in a nuclear power plant accident. The magnitude can be summarized by saying that air pollution, mostly from coal burning, but somewhat from oil burning also, causes more effects on public health than would a Chernobyl-size nuclear accident every year.

But there remains a huge uncertainty which is related to item (iii) above. Although it is likely that air pollution from fossil fuel burning causes adverse effects on health and likely that there is low dose linearity, it is far less sure what aspect of the air pollution is the problem. I believe that fine particles are involved. But exactly what aspect? Amdur [1] showed that sulfate particulates are worse than sulphur dioxide for guinea pigs. But sulfur dioxide, emitted as a gas from power plants and evading all the particulate traps, converts to sulfates in the power plant plume as demonstrated unequivocally in Figure 6 from [23]. The fraction of coarse particles mode decreased and the fraction of fine particles increased as the mobile sampling van moved downwind farther from urban influence. There was more time for a chemical reaction as the power plant plume mixed with the background air, and SO2 was converted to sulfate and NOx to nitrate.

Many countries have therefore been active in reducing sulfur dioxide emissions. Are nitrates as bad as sulfates? If so, we must be extraordinarily careful about automobile and truck exhausts. Is it the vanadium or zinc that is attached to the particles? Once we go beyond the simple exhortation, avoid burning fossil fuels, and trapping all the particles or particle precursors, the recommendations for public health are far less secure.

How can one find out? Critics have argued that mankind spends more time indoors than out of doors and that using the outdoor (area) concentrations as an indicator of exposure is extremely approximate. Gases and for the fine particles now believed to be implicated penetrate indoors as shown in Figure 7 from [21]. But it would be nice to have more direct measurements. An ideal, but very expensive, epidemiological study would involve following more people (perhaps 100,000) in a prospective study, with each person carrying an personal monitor continuously instead of using the external, area, concentrations. But it is unlikely that this uncertainty will be resolved by epidemiology alone. It will need a careful combination of laboratory, animal and epidemiology experiments to elucidate the probable causes. Meanwhile it behooves us to be careful how we burn fossil fuels.

We can, if we wish to spend the money, do a lot about particulate air pollution. At a power plant we can burn natural gas. Natural gas burns quite cleanly. It tends to be free of sulfur and of heavy metals like vanadium and zinc. The temperature can be well controlled so that fewer nitrogen oxides are produced. If we wish to burn coal we can first gasify it and remove these undesirable pollutants. Motor vehicles could also be controlled. Catalysts already do a great deal, and we can demand that SUVs and commercial vehicles are similarly equipped. Hybrid, gasoline/electric engines enable the fuel to be burned more cleanly and when fuel cells are used the particulates are more easily removed. Pure electric engines would put all the pollution at the power plant. BUT the overall simplicity goes down (causing extra expense) and the efficiency may go down, increasing the production of CO2. And, of course, we can always replace the fossil fuels completely by some other fuel.

3. The arsenic problem

(principal reference: the arsenic website:

This workshop includes the problems of heavy metals in the environment. Arsenic is a heavy metal and it has recently been seen as a cause of many problems. Arsenic has been used since 3000 BC (Partington, 1935). It has long been known to be acutely toxic. Arsenic in water at 60 parts per million (ppm) will kill promptly. The limit on arsenic exposure in drinking water was set primarily to be sure to avoid these acute toxic effects. The arsenic limit set by Bangladesh, the United Kingdom, and the United States is currently set at 50 parts per billion (ppb). Until recently this was also the standard that was recommended by World Health Organization (WHO). But WHO lowered their recommendation to 10 parts per billion (10 ppb) and in the June 22nd 2000 Federal Register the US Environmental Protection Agency formally proposed lowering the standard to 5 parts per billion (5 ppb). This has raised a great furor among those concerned with developing countries. What is it all about?