Example of a common case study

The text on this page is taken from an equivalent page of the IEHIAS-project.

Example Common Case Study

Aim of the Common Case Study

In addition to the development and enhancement of single methods one focus of HEIMTSA/INTARESE is to test the integrated environmental health impact assessment system (IEHIAS) developed in INTARESE WP 4.2 and HEIMTSA WP 5.2 and to apply the INTARESE/HEIMTSA methodology to realistic policy scenarios, in order to (i) test the usability of the full chain methodology of IEHIA and the IEHIAS as a whole; (ii) identify any important gaps; (iii) generate results; and (iv) discuss and evaluate both the methodology (for its reliability and completeness) and the results (for their plausibility and practical use).

The so-called Common Case Study is the means to fulfil these objectives. Its aims are (i) to assess environmental health impacts of high-level, cross-cutting policy issues at EU level and (ii) to provide a full example of an integrated environmental health impact assessment according to INTARESE and HEIMTSA recommendation.

Questions

Policies and measures for mitigation of and adaption to climate change are nearly always chosen with a focus on the reduction of CO2-eq and the cost of the measures. However, there may be relevant side benefits or damages, e.g. decreases or increases in health impacts. Those should also be taken into account during the decision process.

To inform decision makers about these side effects, the Common Case Study answered the following questions:

What are the (negative or positive) health impacts of climate change policies in Europe for the years 2020, 2030, and 2050? Specifically,

a      How do EU climate mitigation policies, i.e. policies with the primary purpose of reducing the emissions of greenhouse gases (policies and resulting measures), affect environmental health impacts in Europe, e.g. increased use of biomass as energy source?

b      How do EU climate adaptation options and policies, i.e. policies that reduce negative climate change impacts, affect environmental health impacts in Europe?

Scope

Spatial boundaries:

The case study looked at the European scale (EU29). The spatial resolution differed for sectors and pollutants as appropriate. For air pollutants, e.g. it was based on 50x50 km2 Emep[1] grid cells for regional effects and on a smaller grid for local effects (e.g. traffic in cities). For indoor air pollution parameters were use on a country level including probability distributions. For pesticide modelling a spatially not explicit model that, however, includes trade between EU and non-EU countries was applied.

Temporal boundaries:

The case study looked at a base year (for emission scenario modelling, 2005) and developed scenarios for the future years 2020, 2030 and 2050, i.e. described the state of the system (policies, physical parameters…) and emissions of pollutants in these years. While scenarios for 2020 and 2030 could be developed with a higher degree of certainty, i.e. the possible ranges of parameters were large but still limited; scenarios for 2050 had to be based on less certain assumptions and, thus, bear higher uncertainty. Nevertheless, a quantitative scenario description proved to be helpful in exploring the possible effects of policies including emission mitigation measures.

Effects of emissions might be observed only later than the time of emissions (e.g. exposure is delayed due to a slow dispersion of stressors in the environment; or health impacts can occur only years after exposure) but are attributed to the year of emission – in this case they were discounted to the year of emission to reflect the time preference people give to effects in the future.

Population:

Receptor for the exposure was the European population. According to needs it was stratified by age groups and gender for each 50x50 km2 Emep grid cell. Its growth was also projected to the years 2020, 2030, and 2050 separately for each age group.

Age/sub groups in the PM2.5 exposure modelling were: 0-14, 15-64 and 65+ years of age. The group 15-64 was split by working and non-working people. All groups were stratified by gender.

Age groups for health effect estimation (impact functions) differed from the personal exposure modelling groups. They were 0-1, 0-3, 5-14, 5-16, 0-16, 0-18, 15-64, 18-64, 20+, 25+, 27+, 30+, 65+, 18+ and all ages.

The diversity of subgroups was dealt with in such a way that the age groups of the impact functions were generally used when applying the impact functions to exposure (e.g. concentration or intake fraction). Only for personal exposure modelling to PM2.5 the age groups had to be mapped to fit each other, i.e. age groups were split even finer, the impact functions were applied and the finer age groups aggregated once again to the exposure modelling subgroups.

The reason of working with different age groups during the application of impact functions is that the impact functions were either derived in studies looking at those subgroups (e.g. when looking especially on children) or are only applicable to those subgroups (e.g. susceptibility to ozone for elderly, or infant mortality for infants).

The reason of working with different sub groups during the personal exposure modelling to PM2.5 is to facilitate the comparison of different sub groups, i.e. to explore the impact of different mitigation measures and scenarios on the sub groups.

Stressors:

• Outdoor air pollutants: emissions of primary PM10 and primary PM2.5 including compounds, NO2, NOx, SO2, NMVOCs, NH3

• Indoor air pollutants: PM2.5 and PM10, mould/dampness, CH2O (formaldehyde), environmental tobacco smoke (ETS), radon

• Persistent organic pollutants (POPs): PCB-153 and 2,3,4,7,8-PeCDF

• Noise due to road traffic/transport

• Pesticides: 14 fungicides, 49 herbicides, 7 insecticides, 3 plant growth regulators

• Greenhouse gases (GHGs): CO2, CH4, N2O They are only taken into account for assessing the reduction of GHGs, i.e. how many CO2-equivalents a policy scenario or mitigation measure reduces. No health effects are implied by GHGs.

• Heat: It was explored how measures in urban development like shading impacts on heat exposure of the population.

Health effects:

Impact functions were given as European average. For some functions, due to differences in background rates of disease, impact functions were additionally given for the regions Western, Eastern, Northern and Southern Europe.

• PM10: cardiovascular hospital admissions, respiratory hospital admissions, asthma medication usage (children and adults), lower respiratory symptoms including cough (children and adults)

• PM2.5: mortality, work loss days, minor restricted activity days, restricted activity days

• Ozone: mortality, respiratory hospital admissions, asthma medication usage (children and adults), lower respiratory symptoms excluding cough (children), cough (children), minor restricted activity days

• ETS: coronary heart disease hospitalisation, lung cancer, sudden infant deaths (SIDs), lower respiratory illness symptom days and hospitalisation (children), cough (children), wheeze (children), asthma induction

• Radon: lung cancer

• Formaldehyde: asthma

• Naphthalene: cancer

• Mould/dampness: wheeze (children and adults), asthma development (children and adults)

• Noise: % highly annoyed, % highly sleep disturbed, myocardial infarction

• Pesticides: generic cancer (It is acknowledged that it is likely that other health effects like neurotoxic effects occur, however, no dose response functions could be found.)

• POPs: generic cancer

• Heat: summer mortality

http://www.eea.europa.eu/data-and-maps/data/emep-grids-reprojected-by-eea

Overall methodology

This study follows the approach of integrated environmental health impact assessment (IEHIA) developed in the projects INTARESE and HEIMTSA. An IEHIA is an inclusive and, as far as feasible, comprehensive assessment of the risks to, and impacts on, human health as a result either of exposures to a defined set of environmental hazards or of the effects of policies or other interventions that operate via the ambient or living environment. (Briggs 2008).

The approach taken in this case study is to develop scenarios for a baseline year (2005) and several future years (2020, 2030 and 2050). Within these years a business as usual scenario is analysed (all policies and measures that are adapted and implemented in the respective year). Furthermore, a policy scenario with additional policies is analysed (it includes the climate change policies of interest). As a final step and main result the difference between the two scenarios in each year is explored.

The policy scenario includes all the policies to achieve the 2°C aim by the EU (mainly, but not exclusively, the implementation of the climate and energy package (EC 2008, COM 2008); in addition further policies up to 2050). To assess the importance of single measures within this bundle of all measures, the impact of single measures on emissions (and housing parameters)

References: Briggs (2008). David Briggs: "A framework for integrated environmental health impact assessment of systemic risks." Environmental Health 7(61).

COM (2008). Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. 20 20 by 2020 - Europe's climate change opportunity. COM(2008) 30 final.

EC (2008). Impact Assessment - Document accompanying the Package of Implementation measures for the EU's objectives on climate change and renewable energy for 2020. Commission Staff working document, SEC(2008) 85/3, January 23, 2008, Brussels.

Screening

During the issue framing and scoping phase of the common case study emission mitigation measures in different sectors were coarsely investigated to assess if they should be included into the detailed analysis. During the screening simplified and rough methods were used to estimate the order of magnitude of the health effects avoided or induced by these measures.

For some mitigation measures, the change in health impacts (DALYs) per ton avoided CO2-eq was calculated. A negative figures means, that a policy reduces greenhouse gas emissions as well as health impacts. In monetary terms, a mDALY can be roughly translated into around 40 €, future marginal avoidance costs for CO2 in 2020 could be in the range of about 30-40 € per ton. Thus, a value of 1 mDALY per avoided ton of CO2-eq would mean that the impacts of the mitigation measure on health effects are about as important as their climate change mitigation effects. For other policies, not the absolute damage costs but the changes in health effects between two example scenarios have been estimated (answering the question how would the health effects change if the exposure changed by xy%).

For a smaller number of other measures, a qualitative analysis was conducted, i.e. it was discussed qualitatively, why it is plausible that the use of a measure has a non-negligible health impact.

Some policies could not be quantitatively assessed in the screening process but were identified as less relevant for the detailed assessment and, thus, not further investigated.

Examples for screened mitigation measures in the transport sector are increased energy efficiency, shifts in transport modes, alternative fuels and drive trains, and economic policies like increased fuel tax and city tolls. Screening indicated that the avoided health effects due to these climate mitigation measures range between 0.1 and 1.5 mDALYs per ton avoided CO2-eq. Different stressors like air pollutants (main influence), persistent organic pollutants and heavy metals were included.

Examples for screened mitigation measures in the energy sector are increased energy efficiency (e.g. insulating buildings to reduce heating demand), and changes in electricity supply mix including larger share of bio mass burning, larger share of bio fuels, larger share of renewable energies, and heat and combined heat and power. Screening indicated that the avoided health effects due to these climate mitigation measures range between 0.54 and 1.23 mDALYs per ton avoided CO2-eq. Different stressors like air pollutants (main influence), persistent organic pollutants and heavy metals were included.

The main screened mitigation measure for indoor air quality was change in housing quality, leading to ca. 30% less health effects under certain assumptions like reduced exposure to dampness/mould. Radon was identified as further pollutant to be considered in the detailed analysis.

Health effects due to pesticides via direct application onto plants and further following the food chain were not investigated during the screening process. They were, however, included in the detailed analysis.

Health effects due to pesticides present in drinking water were investigated in the screening process. Due to a serious lack of data and the use of a methodology of limited use screening results were very uncertain and probably overestimating the health effect risks – health effects might even by near zero. The main finding of this screening, therefore, has been to highlight the problems in developing such a model at the European level for drinking water, and to recommend that this particular area of assessment be excluded from the common case study.

Disinfection by-products (DBPs) in drinking water were considered in the screening process. One of the findings was that the changes of DBPs in drinking water were more related to climate change as such than to climate mitigation and adaptation measures. It was acknowledged that the water sector was not well adapted to the pre-determined methodology developed for assessing the main stressors (like air pollution) under the case study framework. Thus, it was decided to exclude DBPs from the detailed analysis but to investigate them separately.

Health effects due to road traffic noise were not investigated in the screening process but were included in the detailed analysis.

Changes in meat demand were investigated as climate mitigation measure in the agricultural sector.

Health effects due to reduction of heat in urban areas (heat island effects), as an example of a climate mitigation measure, were not quantified during the screening process. They were, however, included in the detailed analysis.

Connections and interrelations between the policies (e.g. insulation of buildings and reducing heating and energy demand leads to changes in indoor air quality; higher demand of bio fuels in transport and energy sector influence on the agricultural sector) could not be taken into the account in the screening process. They did, however, play a major role in the detailed analysis of this case study.

Definition of main emission scenarios

Emission Scenario Approach

In this study the effects of climate change mitigation policies and adaptation policies on human health are assessed by simulating the implementation of single policies and measures embedded in scenarios of the future development of the relevant parameters and systems.

For each of the years 2020, 2030 and 2050 two scenarios were defined:

a The Business as usual Scenario (BAU) describes the development of an energy system without any additional energy or climate policies implemented after the year 2012. Further relatively high fuel consumption of newly registered cars was assumed.

b The Climate Policy Scenario describes a prosperous world, where the world together manages to reduce greenhouse gas emissions to an amount that leads to an average temperature increase of not more than 2°C (with a 50% propability). This means that the CO2 concentration has to stabilise on a level of 450 ppm. Consequently, annual reductions of GHG emissions of 71% in 2050 compared to the year 1990 need to be achieved. This reduction includes the 20% GHG reduction target for 2020 compared to Kyoto base set by the EU-27. Additionally, the EU goal of a 20% share of renewables of final energy consumption in 2020 is implemented. This includes a reduction of emissions of 21% in 2020 compared to 1990 in the sectors covered by the emission trading system. In the transport sector, emission standard EURO VI is to be introduced in 2014.

2005 (2010, 2000)	2020 BAU	2030 BAU	2050 BAU
	2020 Climate policy	2030 Climate policy	2050 Climate policy

+ additional scenarios representing single mitigation measures

Political frame

In 2007, the EU agreed on an independent commitment to achieve energy and climate targets for (EC 2008)

• a at least 20% reduction of green house gases by 2020 compared to 1990 levels and

• a 20% mandatory use of renewable energy by 2020 including a 10% share of bio fuels in petrol and diesel.

Along with that, the “Climate action and renewable energy package” describes the contribution expected from each Member State and proposes policies to achieve them (COM 2008):

• strengthening and expanding the EU Emission Trading System (ETS)

• setting emission targets per Member State for non-ETS sectors (transport except aviation (included in ETS from 2012 on), housing, agriculture and waste) ranging from -20% to +20% depending on their wealth

• national renewable energy targets with a minimum share of 10% bio fuels in petrol and diesel by 2020

• development and safe use of Carbon Capture and Storage (CCS) and

• increasing energy efficiency to save 20% of energy consumption by 2020.

Consistency and interrelation of policies in different sectors

During the development of the scenarios consistency of assumptions from different sectors and models was striven for as far as possible. The interrelation of the different policies in the different sectors was especially minded.

Increased use of bio mass and bio fuels in the transport and energy sector leads to increased cultivation of energy crops, which again leads to a competition of demand between agricultural land for energy crop cultivation and cultivation of crops for food consumption. Increased bio mass burning in houses might lead to higher exposure to combustion products.

Further changes in drive trains of vehicles like electric cars lead to changes in the energy demand and, thus, energy supply, which again impacts on outdoor air pollutant emissions.

Insulation of buildings leads on the one hand to energy efficiency and reduced heating demand. This reduces emissions of outdoor air pollutants from power generation. Secondly, however, it may lead to a reduced air exchange rate if the building envelopes are made tighter but a hygienic air exchange rate is not ensured. In this case pollutants can accumulate indoors causing a antagonistic effect to the one intended.

Considered pollutants and sectors

The emission scenarios were developed for the following years: 2020, 2030 and 2050. As a base year 2005 was chosen. The scenarios are based on an emission data base, which was especially developed for the common case study. The data base includes emission data for

• air pollutants: NH3, NOx, PM10, PM2.5, PMTSP, SO2, NMVOC and CO and

• greenhouse gases: CH4, CO2 and N2O.

The data base covers all anthropogenic sources (energy, industrial processes, product use, agriculture, waste and other sources). It includes emissions data for 29 European countries (EU 27 plus Norway and Switzerland).

The considered sectors are shown in Table 1.

NRF Code	Source Category
1.A.-1.B.	Energy (Energy Industries, Manufacturing Industry and Construction, Transport, fugitive emissions from fuels)
2.A.-2.G.	Industrial Processes (Mineral, chemical, metal and other production industry, other)
4.A.-4.G	Agriculture (Manure management, agricultural soils, field burning, other)
6.A.-6.D.	Waste (Solid waste disposal on land, waste water handling, waste incineration, other waste)
5. and 7	Other sources

Example Common Case Study: Estimating Health Impacts

Within the HEIMTSA-INTARESE full chain methodology, estimation of health impacts follows assessment of exposure scenarios – baseline, and alternative taking account of the policies being evaluated. Methods for linking exposures with health differ according to whether risk estimates come from epidemiology or toxicology. The Common Case Study (CCS) includes numerous environmental risk factors, including outdoor air, indoor air, noise, pesticides and POPs, and so it uses both approaches. For most of these pollutants causal relationships with health are established; at issue is the size of the policy-attributable health effects in the EU-wide target population of the CCS.

This short article summarises how the epidemiological approach was implemented in the CCS, with special reference to the most influential pathway: the effects on mortality of long-term exposure to fine respirable particles (PM 2.5), which has particular methodological complications.

One key element is the exposure-(or concentration-)response relationship (ERF), usually expressed as % change per unit exposure in the risk of disease or other health effect (e.g. minor physical symptoms, hospital admissions, death). ERFs used in the CCS followed detailed epidemiological review, often including meta-analysis of results across several studies, and drawing on evaluations of established expert groups. Uncertainty was typically expressed via confidence intervals, but sometimes included expert elicitation to capture, for example, transferability of ERFs between populations. Sometimes conversion factors were needed so that exposure estimation and ERFs used the same exposure metric.

For mortality and annual average PM2.5 both HEIMTSA (detailed epidemiological review) and INTARESE (review and meta-analysis) independently endorsed a risk estimate of 6% increase in mortality hazard (95% CI 2-11%) per 10µg/m3 PM2.5 previously used in the EU CAFE and other policy applications.

Sources of background rates (incidence or prevalence) across the EU for the identified health outcomes were reviewed. Routinely collected mortality data by age and sex were easily available; background rates for most other health effects required extrapolation from individual countries or specific studies. Care and judgement were needed to ensure that ERFs and background rates used consistent definitions of the same health effect.

ERFs and background rates were linked to provide a set of impact functions, typically expressed as the number of additional cases per year per unit exposure per 100,000 population. Sometimes this was simple; often some processing and judgement were needed, for consistency in time resolution or sub-population, including adjusting background rates to apply to the unexposed population only.

Mortality from long-term exposure to PM2.5 presented particular difficulties, because changes in mortality risks affect population size and age distribution over time. Using life table methods we estimated the long-term impact, in life-years, of a 1-year reduction in 2010 of 1µg/m3 annual average PM2.5. Analyses in England and Wales, Italy and Sweden gave consistent results over countries and sexes; a weighted average of 235 life years per 1µg/m3 PM2.5 per 100,000 population aged 30+ was used for Western Europe. Impacts for Poland were somewhat higher, with a corresponding value of 312 life years used for Eastern European countries.

Example Common Case Study: Monetisation of Health Outcomes

The full-chain approach to health impact assessment that has been used in HEIMTSA allows for the aggregation, in monetary terms, of the disparate health end-points resulting from pollution. Use of the money metric is designed to capture people’s personal preferences in relation to the health end-points. Thus, the common measure is of the individual’s willingness to pay (WTP) to avoid a specific health condition. Whilst there is a pre-existing body of work the breadth of the coverage of environmental media in HEIMTSA has necessitated both a re-evaluation of existing unit value estimates, and an expansion of the number of end-point unit values required.

Table 1 presents a summary of the unit values derived in the course of the HEIMTSA project. These values are the result of both an evaluation of the evidence available in the existing literature and new empirical research undertaken in the project. For each health end-point, the unit values are identified on the basis of an informal meta-analysis of the evidence, accounting for the distribution of available values and an assessment of the quality and geographical focus of each study. Thus, studies whose results converge on modal values, which use state-of-the-art non-market valuation techniques, and that are undertaken within the EU are given greater weight in determining therange of values for each health end-point. It is obvious from the values presented in Table 1, and specifically the range of values associated with a number of the end-points, that there is considerable uncertainty in health valuation. For example, for valuation of life-years and neuro-developmental disorders, there is a difference between low and high estimates of factors of 5 and 8, respectively. In cases such as for anaemia, there is insufficient evidence even to provide a range. The uncertainty derives from a combination of the paucity of the evidence base, the difficulty that people have with identifying their preferences for (avoidance of) health conditions, and the lack of maturity in the study methods themselves. Thus, in the instance of valuing avoidance of the risks of premature death, methods are only now being developed that successfully communicate to survey respondents the small changes in risk that tighter environmental regulation would result in. For example, computer graphics allow such information to be presented in a variety of ways, often with an accompanying voice-over.

The uncertainty presents a challenge in the use of such values though sophisticated modelling approaches such as Monte Carlo Analysis lend themselves to the treatment of this uncertainty and are likely to be equally necessary in dealing with the uncertainty attendant in preceding stages of the full-chain approach. It is also worth emphasising that the HEIMTSA project has afforded the first attempt at monetising a more complete range of health impacts. In so doing, it has served to establish a baseline for health impact assessment to work with from now, and highlights where the gaps are for future empirical research.

Table 1: Summary of Monetary Values relating to specific health end-points (€, 2010)

(Table not found)

References:

Briggs (2008). David Briggs: "A framework for integrated environmental health impact assessment of systemic risks." Environmental Health 7(61).