You are on page 1of 46
AIR MONITORING PROGRAMME DESIGN FOR URBAN AND INDUSTRIAL AREAS Prepared by the Secretariats of the WORLD WORLD HEALTH and the METEOROLOGICAL ORGANIZATION ORGANIZATION following a WHO/WMO Consultation on Air Quality Monitoring in Urban and Industrial Areas, Geneva, 1976 WORLD HEALTH ORGANIZATION GENEVA 1977 WHO Ofiset Publication No. 33 WHO offset publications are intended to make generally available material that for economic, technical, or other reasons cannot be included in WHO's regular publications programme and would otherwise receive only limited distribution. They are usually reproduced by photo-offset from typescript, rather than by letterpress, and do not necessarily receive such detailed editorial revision as other WHO publications. ISBN 924 170033 5 © Wortd Health Organization 197 Publications of the World Health Organization enjoy copyright protection in accordance with the provisions of Protocol 2 of the Universal Copyright Convention, For rights of reproduction cor translation of WHO publication, in part or in tovo, application should be made fo the Otlice of Publication, World Health Orgaoiction, Geneva, Swtzcrand, “The World Health Ortan~ ization welcomes such applications. “The designations employed and the presentation of the material in this publicstion do not imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health, Organization concerning the legal status of any counlry, territory, ity oF area oF OF its authorities, of concerning the delimitation of its frontiers or boundaries. ‘The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of ‘similar nature that are not mentioned. Errors and omisions excepted, the names of proprietary products ate distinguished by inital capital Tetters. PRINTED IN SWITZERLAND Frese ee nett eee Tatroduction TTT EPEDPIDIIIint Chapter 1. Monitoring objectives and approaches... 2... sL Objectives ee eee eee 112 Approaches to monitoring programe development | 1! Chapter 2. Monitoring network design... ee eee eee 2 ee Chapter a 3 3 1 Background information needed oe ee ee eee 2.2.1 Sources and emissions oo... eee eee 211.2 Wealth and demographic inforeation | |.) |] 2.1.3 Meteorological information...) 1.1. 21114 Topographical information | > er 2.1.5 previous atr quality taforeation |). |) | 2.1.6 Land-use zoning considerations «2.1.1. 2 Organization and resource requirements so... vv oe 2.2.1 Organizational arrangenents . 2... 2.2.2 Personnel, equipment and finance» 1.) 3 Design considerations oe eee 2.3.1 Selection of pollutants cee ee 2.3.2 Selection of the area co be siudiea ||) | | 213.3 Number and distribution of sampling cites |. | 213.4 Sampling duration and frequency. == > - . 3. Stacion location eriterta . sees 1 epresentativeness 6 ee 2 Requirenents for comparability | 11111111 ly 3. Physical requirements... 00st sss Chapter 4. Other essential components of the monitoring programe 4.1 Laboratory requirements oe ee eae Red (esslity jessie eae eee er 413 Data acquisition and handling procedures |) 2111! 44 Date presencacion Dever ct liiiill iadeconceph eee et yr eye tae tee Annex 1, Principal sources and pollutants of possible concern in irre eerie ner Annex 2. Rxamples to {liustrate the guidelines gor the develop TRent of an air pollution monitoring programe -- ss. + + Eximple 1. Starting ca air pollution aoniéazing prow fsranme based primarily on manually operated inetriments . . Example 2, Development of an urban air pollution aonitoring progranne based primarily on sutomated equipment vs ss Example 3. Developing an air quality monitoring programe in the vieinity of a large single source wv ts ne es Annex 3. List of participants in the WHO/WMO Conseltation On air {Yuality Monitoring in Urban and Industrial Areas, Geneva, Be23 Fone IS ee ee ee tee eect eee tees 28 28 28 2 20 2 35 35 38 a 45 PREFACE oth the World Health Organization (WHO) and the World Meteorological Organization (WO) have for many years been engaged in technical coopera tion with Meaber States in establishing or developing environmental control programmes, and geveral projects have been undertaken jointly by the tno Organizations.1»2, This collaboration was strengthened ven the Global Eqvironmental Monitoring Systes (GEMS) was established as part of the Daited Nations Environment Programe (UNEP). ‘The World Meteorological Organtzation has organized a global back- ‘ground air pollution monitoring network operated by Member States in rural Gnd renote locations. As air pollution is one of the factors chat ‘sffece human healeh, che World Health Organization begap to monitor air quelity in selected urban and industrial areas in 1972.” noth WMO and WHO programmes are part of GEMS and are operated in cooperation with and with’ the support of UNEP, A group of experts was convened in Geneva in June 1976, cosponsored by both Organizations, to consider the principles for the design of urbai alr monitoring networks, including aeteorological aspects. The participants (see Annex’ 3) reviewed @ document that had been drafted prior to the meeting on the basis of working papers propared by Mr G. akland, Dr D. Szepesi and Mr R. Waller. The document was eubsequently edited for this publication by # saall panel in collaboration with WMO and WHO staff. ‘The wenbere of the panel vere Dr J. Mahoney and Mr R. Waller, ho had bork also participated in the consultation, and Dr D. Shearer, WHO Collaborating. Centre on Environmental Pollation Control, United States Environmental Pro~ tection Agency, Washington, DC, USA. InrRoDucTION ‘The sain purpose of this publication {a to provide guidance on the gn of air quality monitoring programees for urban and industrial areas Such programmes are needed for alaost all the actions taken to prevent for abate air pollution, from the initial assessment of existing conditions, to the enforcenent of current control regulations, to the evaluation of the effectiveness of abatenent programes, and finally to the development of nee control measures, Tt is alao hoped that this publication will contribute to aore effective hamonization of air monitoring procedures, permitting the Comparison of date obtained throughout the world, ‘The tnformation contained in this publication attenpts to provide general guidance on shen the institution of monitoring programes should be considered, what should be monitored, where monitoring should be carried out and hoy it should be conducted, Air pollution control procedures are not discussed, although they are dotinively interrelated with those of air quilicy monitoring, Methods of analysis of air pollutants have bees presented elsexhere.?~10 ‘me information given should also serve to guide the further deyplop= ‘nent of the Global Environmentel Monitoring Systen (GHYS) activities in urban areas, In addition, emphasis 1s placed on the need for @ full land continuing cooperation between sir pollution control agencies vhick fare involved in sir quality apaltoring and the local and national meteoro~ logical eervices. ‘Air quality onitoring proposals should alvays be viewed in the broad context of environmental maaagenent, The decision-maker must often assign priorities amongst a sultitude of environmental concerns and he must Gocide whether an air qualicy ronttoring programme is justified, and if fao, what level of technical support it thocld receive, Tt is important to beware of establishing too sabitious 2 programe unless there is a clear understanding of how the air quality data will be ust ‘this publication ie prisarily directed at adninisteative and echnical personnel in those countries that are becoming concerned about air pollution aspects of an increasingly developing industrial economy. The Goldance provided should also be useful to those countries that have al= Feady established 4 monitoring programe and are aaking decisions for ferther development of thely monitoring network, In this connexion it should also be noted chat industry should be encouraged to monitor nat only their plant enissions, but also che ambient air in and around their festablishents, The guidelines presented will also aid them in this An overall picture of the major steps for the development of an air wonitoring programe is presented in Fig. 1, The various developmental stages are show in chronological sequence; some steps include « number of activities that can be carried out aimiltaneously. Te toxt 1a organized into four chapters as follows; (1) monitoring objectives and approaches, (2) monitoring network design, (3) station loca: tion criteria, and (4) other essential components of the monitoring pro- gramme, The principal information is contained in Chapters 2 end 3, shere specific guidance is provided on the design of air quality monitoring networks. Chapter 1 serves primarily ax an introduction to these chap- ters and Chapter 4 gives some additional helpful information. In addition, three examples are given in Annex 2 ¢0 illustrate the dovelopment of different types of air quality monitoring programnes, The First example shows the developuent of @ programe based primarily on the lise of manually operated instruments; the second is based on the vse of autonatic monitoring iastrusentation; snd the third deals wich the development of & monitoring programe in the vicinity of a large single source of air pollution, It shovld be stressed that these examples should only be taken as general guidance to illustrate some of the types Of prograames that could be developed for a particular need. Specific Suggestions about ninbers of samples and costs are included; however, All such decisions and calevlations mist be made locally in any real ‘situation, folloving the guidelines set out in the main part of the cext. FIG, 1 STEPS IN THE DEVELOPMENT OF AN AIR QUALITY MONITORING PROGRAMME STEPS Selet monitoring approach Existing | = «i peters EH eee’ |") capigmmee | '[_ Sater Bae pope Eom (1) Personnel (2) Funds (G) Equipment (1 Poliatants to bbe monitored (2) Number and type of system (2) Genera types of iratruments (4) Expected tres of data (5) Plans tor fuality ‘Sauranee and parading of ‘monitoring rotwork STEP 6 Design network (1) Number and Semiunon otrione 121 Support portions (21 deat (6) Bataan procedures: a relat 1) cao ia a ioe sreP7 Implement programme (1) Acauite sites (2) Purchase (a) initio vaining HAPTER 1, MONITORING OBJECTIVES AND APPROACHES 11 objectives Before considering the establishment or expansion of an air quality sonitoring programme, it 1s essential to examine the objectives care~ fully 4f the appropriate data are co be collected with a mininun of effort and cost. Although it may be tampting to design a system that could serve ‘a multitude of different objectives and associated data needs, in practice ix appeare that only certain combinations of objectives are realizable with a given network, For example, it is generally not possible to wee a net~ work designed to monitor long-term trends of air pollution levels to inves~ tigate a specific complaint, OF course, it is possible to modify a net~ work designed primarily for one purpose go that it will serve another a wells Monitoring objectives have been divided into two groupe, The first group (a and b below and in Table 1) includes the objectives for a gonitor- ing progeamne for an average industrialized efty with an actual or potential air pollution probles, They cover the besie monitoring requirements and fnost of the rensining chapters deal with further inforeacion on these. The second group (cnL below and in Table 1) includes wore specializes monitoring objectives, which are generally more optional in nature (and also technically somevhat more complex) and, according to the local sitva~ tion, say of may not form part of the basic’ programe. ‘the-basie monitoring programe will provide the essential data re- quired to develop sir quality standards and will in general permit the develoment of viable air pollution control programe, Normally the faajor air pollutants (see Annex 1), such as suspended particulate matcer, sulfer dioxide, carbon monoxide, nitrogen oxides and oxidants, would be considered for inclusion in chia type of monitoring operation, — Depending fon the type of industries active in the general area, other pollutants, for exanple, hydrogen sulfide, may be added.!? “The meteorological pae- eters to be measured would be wind speed and direction, If possible, the frequency of inversions should also be recorded, “The objectives in the first group are as follows: ) To observe long-term trends ‘This is to detect any deterioration in air quality arising from residential or industrial development. For chis purpose alone a network using manually operated equipment would be sufficient; 2-hour samples could be collected intermittently over relatively long periods and meature~ ents might be made at only three or four sites within a given urban ar (with the important proviso chat these should reoain in operation at the sane places for many years), The aiting mat, hovever, be considered very carefully Co avoid undue influence from nearby sources. Much additional information can be guined, and the network can bemade wore versatile, if uninterrupted observations are aade over 24-hour periods. Statistical analyses can then be done to provide not only the yearly averages, but the Frequency distributions of the daily values, Trends can be atudied separately for weckdaye versus veokends, and interrelationships with meteorological factors can be exanined, Because of the large seaonal variations in the concentrations of some pollutants, it is always important when assessing tronds to make sure that sampling has been carried out throughout each year, or at Least for periods representative of it, Major changes in the weather pattern from One year to another can affect even the annual weans and, to observe under~ Tying trends, data for at least five consecutive years my be required, (b) To judge compliance with air quality standards and to evaluate control _atretegies In general, insufficient attention has beon given to the monitoring procedures to be adopted when recomaending air quality standards, It te Important for thea to be specified in detail, since reaults depend so much on the exact location of saaplers and on the averaging periods adopted, For anbient air quality standards the aim should be to assess concentra~ tions that are relevant to the exposure of the general population, vith- out undue bias from local sources or unusual dispersion conditions, — When fnew control measures are introduced, a long series of resulta, extending over at least five years, is required to establish their effectiveness, fand the monitoring network chould, if poscible, be set up before the con trols come into operation, Usually 24-hour averages are required, In sone cases, such as for carbon sosozide, averages of less than 24 hours are needed, ‘he second group of objectives includes the folloving: (c) To activate emergency control procedures ‘This is primarily a problem of forecasting the weather and, in turn, the Likely concentrations of key pollutants for periods of about 26 hours ahead. A prerequisite is that nathenstical models be developed relating Concentrations of air pollutants co meteorological variables. — Forecasts of pollution levels derived from such models need to be backed up by Foutine observations in the urban area, using sampling periods of the order of one hour. Systens designed specifically for energency control can be highly sophiaticated, with facilities for giving advance warnings, but these are only justified Sf auch warnings can be used effectively to protect the population. (2) Te evaluate risk to husan health ‘the affects of urban air pollutants can be considered under tvo head~ ings: acute effects, usually observed in terms of day-to-day changes in some index of health; and chronic effects, which gradually become manifest after many years of exposure to pollution, For the first of these it is essential co have measurenents of pollutants over short inter- vals; 24-hour sampling periods are in general adequate but to assess exposures to peak concentrations during the day observations averaged over shorter periods may be required, in relation to chronic effects, ‘annual averages may suffice but, since a knowledge of seasonal variations ind peak values eay be relevant, series of observations over 24-hour periods, as required for acute effects, provide a suitable basis. Wnere concentrations vary greatly over distances of @ few metres, a chey are lable to in the case of carbon monoxide or Lead from traffic sources, chen specially designed surveys are generally required for each study, possibly involving the wee of personal samplers, 10 (9) Te assets other environment risks Risks of danage to plants and trees within the urban area can usually be assessed op the basis of observations over 2échour perlods, as in objectives (3) or (g) above, but since a brief exposure to high concentra Eions (e.g+, of sulfur dioxide or orone) can damage sensitive flovers, continuous monitoring instruments may be required if such considerations are of any great importance, Risks of damage to stonework and other bullding saterials, particular= ly from sulfur dioxide, are best assessed with static monitors, for example, cad peroxide candles, which are exposed to the atmosphere for periods of fone month at a time, Analysis of precipitation samples, especially for acidity, 1s also useful. (2) Tp provide the data bese for land-use planning land-use patterns and their acconpenying activities deteraine to a large extent the types and amounts of pollutants generated in an urban area. With increasing concern for environmental quality, there is s need to T+ Inte existing pollution Levels to the impact of new or expanded urban industrial complexes and of the establishment of new major public works, For sulfur dioxide and suspended particulate macter the instrumenta~ tion and data requirenonts are siailar to those in objectives (a) and (d) above, using instrimencs urually operating for 2é-hour periods, The aet~ work should include sites representative of each kind of land use, such 6 high-density end low-density residential areas, industrial couplexes and comercial districts, IMO regional stations ney provide additional use~ ful dats (g) To validate dispersion aodels Disperston models may be related to the emissions fron a single source of to the integrated effects of sultiple sources in a large com munity and monitoring requirenents will vary accordingly, Ta determin ing the frequency of sampling, the nature of the model mist also be con- sidered, whether it 1s to be used in relation to short-tem variations in pollution over periods of 24 hours, and down to a fev minutes, or in the ‘assessnent of Long-tera averages (1 month ~ 1 year), In most cages an extensive monitoring network using eutomacic instruments is required for 4 period of about one year, to “calibrate” the model. (h) fo dnvestigate specific complaints For the investigation of specific complaints the sampling may, in contrast vith that in earlier sections, be deliberately aimed at picking up pollueion fea local sources, Since monitoring results are influenced by musi changes in wind direction, a large aber of canplere may be required ‘and continuous monitoring equipaent may have to be included so that tran~ Stent peaks can de observed, There may be difficulties if the type of pollueant cannot be specified in advance, for it would probably not be practicable to eet up special sampling stations for a variety of pollutants, Mobile sampling stations can be valuable in such cases, (4) Te carry out iniesal assessment surveys ‘here no meagurenents have been made before or vhere a new type of pollution problem is emerging, an exploratory survey may be carried out n TABLE 1 x (ATR QUALITY IRSTRONDUTATION TH RELATION TO ORTECEIVES routaxr onreerives nstanese ee coments sulfur a Avcomutie instruments Continuous High Moderate- Instruments tnclode whole analytical dioxide High process; skilled personnel aust be available to ensure proper fnetioning. aubsdyesfshst — Machanized bubbles 1-248 moderate Moderate Sunple analysis carried out in the laboratory; often used in conjune~ tion with enoke saaplers, auPedveshst Absorption esber womin ow Moderate Applicabte to intensive surveys of stban areas; sanpling at many sites Wisited according to statistically designed schedsle. avert etd peroxide candles month Low tw Ugefol co indicate relative disteibs Hon around major sources; do not give absolute concentrations, Suspended aybydseyE,0,4 Soke samplers 2h Moderate Low operated at a low flow rate, sampling particalate only saail percicles; sone chenicel matter analysis of filter deposit possible ange of chenteal analyses on filter epost posstbie, an Automatic instruments Copttmious High Moderate- operated at lov Flow rates, vith (4 aapoe) Righ—souples collected on filter tapers sone chenical analysis of filter deposit possible Deposited ashyt Dustfall gauges Tnonth Low low Sample large particles only; results particulate in general not relevant to effects on etter health; chemical analysis of deposits possible. carbon au Automatic inetrumente continuove igh Moderate- In addition to on-eite messurenents, monoxide high spot samples may be collected in the field and broughe back co the laboratory for evaluation. Oxides of —aybydye,f,h,i Mechanized bubblers 1th b Moderate Moderate Sample analysts carefed out in the nitrogen, laboratory. oxidaace a Automatic instruments Continuous High Moderate- Instruments include whole analytical high process; skilled personnel must be available to ensure proper func tioning. Oxidants Rak Mechanized bubblers 30min Moderate ‘Moderate Sanple analysis must be carried out (as ozone) in a Taboratory 30-60 ain after sampling. Au Automatic inetrumente Continvows igh Noderate- Instruments include whole analytical, high process; skilled personnel must be available to engure proper func a trend analy evaluate control strategies} ‘Gvaluate Tiek eo human health ‘evaluate isk of environmental damages data base for land-use planning validate dispersion nodels; Investigate complaints initisl assesanent. for a Mmited period using manual samplers or, if available, a mobile sanpling station. . A knowledge of the kinds of fuel used andof industrial activities in the area ds iaportant as a guide to the types of pollucante to be sampled, Where there are requirenents for heating dering part of the year only, or where there are major changes in meteorological factors, the survey should include the period when air pollution Levels are expecte: to be high. 1.2 Approaches to monitoring programme development ‘The design of a monitoring programe will alvays involve che con- sideration of a aumber of broad questions of what, where, why and how to monitor, which must be resolved before the detatled design of the specific network can proceed-!? For example, every moderate or large-scale nonitoring network adheres to one of tw batie approaches: (1) many st Cons with measurements made for one or two pollutants only in each eta- tion; or (2) few stations with meseurenents sade for severe! pollutants in each station, Even this basic division in approach has several other dimensions: general choices of instrument types, plans for the use af the aca, etc. The List below sumarizes the major factore to be considered in the devologment of the approach to a monitoring systen designe (QD Pollutante to be measured, Some networks will include im struments for measuring only the ost common urban pollutants, sulfur dioxide and suspended particulate matter, Other networks will lo in ‘clude instruaente for the other common urban pollutants, as mentioned in Annex 1, (2) Number of stations, This is closely related to the objectives of the monitoring network and the resources availabl (3) Station deployment, The monitoring stations may be at fixed locations, or @ mobile monitoring station may make measurements at & nuaber of different locations, In some situations a combination of fixe ‘and mobile atations {s best, (4) Inetrument types, A brief review of che types of equipment available, as they relate to the monitoring objectives previously described, together with an estimate of relative costs is given in Table 1, This i also deale wich to greater detail in reference 16, (5) Sampling periods. Three principal sampling periods should be considered: shore term (e,g-, 1 hour), moderate term (1 day) and long term (Lmonth), Instruneats gonitoring continuously can provide data for all tine scales, The sampling period is particularly important 1f the monitoring feovlts are to Se compared with standards or criterie, (6) Meteorological information, The availability of the meteora~ logical information requived to locate che monitoring station and to snterpret the sir monitoring data should be deterained, The types of meteorological inatrunents typically required to meet the various sonitor— Sng objectives are given in Table 2. % METEOROLOGICAL, REQUIREMENTS IN RELATION TO OBJECTIVES, Monitoring objective Instruments and measurements Initial arsesement ‘rend analyoie, evaluate risk to human health, evaluate risk of environmental damage Evaluate control strategies Validate depersion modele Activate episode controls Investigate complaints Data base for Land-vse planning Recording wind instruments. Recording wind instruments, thermo- hygrographs, precipitation gauges. Recording wind instruments, thermo- hygrographs and precipitation gauges; vertical temperature pro- file observations useful, often based on radio-sond measurements From local meteorological services Observations from a group of meteorological stations in the area} one site should be a tover station With Camperature and wind observa- tions at different heights; 1 not possible, atmospheric stability Conditions can be estimated by lising ground-based paranetera: radiation intensity, cloudiness and wind observations; consultation with meteorological experts recom- ended, As above; collaboration with local meteorological services essential; immediate access £0 current observe-| tions of vind and tenperature (stability) parame: es required. Portable vind instrunente, thermo eters and precipitation gauges Portable wind atations useful ££ the topography of the area ie cot plicated; vertical temperature observations useful. 1s HAPTER 2. MONITORING NETWORK DESIGN 2.1 Background information needed 2141 Sources and emissions fone of the first steps in the design of 4 monitoring network is to gather information concerning the sources and emissions of air pollution En the rea to assemble an enlscion nventorys The main sources in a city usually include industries, motor vehicles, pover plants, incinerators and heating equipment (see Annex 1). Information should be collected about their mmber, type, size and location, and this should be supplemented with data on types and quantities of fuels used and thetr composition (sulfur, ash and trace-elenent content). In some cases there may be publications that give local, regional and national totals Fuels for stationary sources shovld be considered eoparately from fucks used im transportation. ‘Tere are guides available for making such emission inventories. +16 Once this Lnforaacion has been collected, a rough estimate of the anbtent air pollution concentrations to be expected can be made using dispersion Calculation models.17+18,19" In thie way, an initial assessment of the problems co be expected can be gained in @ short period and at a relatively low cost. then considering the distribution of sources 1t ie important to make a distinction between large sources, which often eait through high atacks, and ‘snali sources, which emit at a relatively low height. Thus, smaller sources ‘nay have proportionately much greater impact on ground Level concentrations fn the gutvounding area thaa the large industrial sources. In this con nexion it ia alao important to make a distinction between pollutants eaitted directly from the sources (primary pollutants) and pollutants resulting from Photochemical reactions (secondary pollutants). The latter are particularly Important vith respect to emiscions from transportation and the petrochenical industry. Since secondary pollutants are formed in the atmosphere the highest concentration may be found at some distance from che air pollution sources and thia aspect ahould be considered when designing the air monitor- ing network. 2.1.2 Health and dewographie information he decision to start a monitoring prograsne, is frequently based upon complaints, In many cages these are caused by nuisances, Like malodours oF dust, The origin and geographical distribution of the complaints, their type and number can help in the design of the sonitoring network. Te is often also useful to collect information on the pollutant danage to plants, animale and materials in the area, a5 an aid in selecting monitoring eites, Information on the distribution of the population within an ares Ls re quired, parcicularly where the objective ie to evaluate human exposure. Ar quality monitoring for epideaiologieal studies is usually done in a number of residential areas having significantly different aiz pollution levels, Information on age and socioeconomic conditions of the population, nornaliy obtained from the resuics of national population censuses, would also be required for these areas. 6 2.1.3 Metesrologiea! tnfornstion contact should be made at an early stage with the meteorological service to learn about the types of information available. Meteoro- logical data are generally assembled for purposes other than air pollution onitoring, e.g.) weather forecasting, air traffic assistance, and agri- ‘culture and hydzological services. Woreover, meteorological measurenents fre not waually made within Large cities. The collaboration of the eteorological service is therefore essential to complenent the existing {information by stationing sone additional instruments in places that are part of the air pollution monitoring programme. ‘The local meteorological services weutLly have general inforastion about climatic conditions in the area, Wind direction, wind speed and temperature variations with the tine of day and year are some of the more comion parameters messured, Other measurements that are often available fre precipitation data, hours of sunshine, relative and absolute huatdity find the potential for fog formation, Tenperature gracient observations fand data on the height of inversion base are very useful but are not 2.1.4 Topographteal information Topography plays an {portant role in the selection of nonitoring sites because of it effect on local wind and stability conditions Many industrial areas have developed in river valleys, where there te an increased tendency for tenperature inversions to develop and to trap the ‘air pollution. In cities built on hilly ground there are substantial variations in concentrations within the urban area. In general, the more complex the terrain, the more sanplers vill be needed to deternine the Gisteibution of pollution. Other topographic features that affect the Aispersion of pollutants include aountains, lakes ané oceans. 21,5 Previous atx quality informsefon ven Lf a continuous monitoring programe has not yet been established, there 1s often some information on air quelity that has been collected in a sporadic manner, for example, special studies done by health and meteorological services, university and scientific researchers or even students preparing their graduation theses. ALL this information should be collected and if posatble, tabulated. Sometimes a first estinate of the magnitude of the problen can be obtained in this way. Caution should be exercised in the use of there data since a variety of sampling and analytical procedures may have bean used. 2.1.6 Land-use zoning considerations Ef air pollution data are to be used in estimating concentrations to bbe expected elsewhere in similar circumstances, it ia important to classify each aite in terms of sources and activities in the area, This fan be done conveniently in terms of the description of existing land use. Generally it is sufficient to esseas this by on-site observation, but in sone azens planning authorities aay have dram up maps shoving land use, In the United Kingdom National Survey, for example, the immediate vicinity of each measuring site vas classified es residential, comercial, indus trial or mixed, qualified by other features such as the density of the housing, and whether there vere parks or other cpen spaces near by. uv 2,2 Organization and resource requirements 2.2.1 Organtzational arrangements The type of organization responsible for the onitoring of sir quality varies from country to country and within countries as vell. In some, responsibility has been assigned to specialized agencies for environmental pollution control, in others, to the health ministries, to other govera- smencal agencies oF co industrial, scientific or educational institutions. Whatever the case, there is 4 real need for the responsible agency to involve other organizations, as appropriate, in the plansing, design and operation of the network. Other orgeniestions that can assist in the davelopment of the monitor ing programe include industrial associations, census buresux and urban planning agencies. Much of the data on emissions, population density and distributions, source locations, traffic distribution, zoning and the ike ean be obtained from these organizations. It is important to ensure seteorological support, parcicularly in the planning stages. The capabilities and experience of other environmental monitoring programmes should also be taken into account. Thue, those responsible for water quality and environmental radiation monitoring programmes as vell as occupational exposure surveillance programmes should be consulted and their facilities used as much as possible, especially for training, Laboratory services and data evaluation. Depending on the air pollution dispersion characteristics (related to topography and meteorology) of the area to be surveyed, the develop- ment of the monitoring network aay require the establishment of cooperative programnes snong communities heving their own jurisdictional responsibilities for environmental pollution control, In such easee dt might be useful to establish some kind of formal cooperative arrengenent, such as a con~ soltative or advisory comitees. In mony cases there are also advantages to be gained by participar- ing 4a international aonttoring programs, auch as those carried out by WHO and WHO in colleboration with the UNEP’ Global Environmental Monitor ing Syeten, These programes include, for exanple, quality assurance components chat could assiat local eir monitoring agencies in the develop= ment of their efforce in this area, 2.2.2 Personnel, equipment and finance Although developaent of an alr quality monitoring programe can be Gitticult, it is posaible co get good prelininary resslts wich simple equiment and 2 restricted budget. The first step shoulé be the selec~ Efon of a small group fo undertake the initial work, One professional, acting part-time or full-time as head of the group, could, with the hel? of one or two technicians, start assembling the available inforeation for the preparation of s report on the conditions within the city. Information on the levels of suspended particulate matter end sulfur Gloxide can be collected vith inexpensive equipment, easily serviced by Seninek{lled personnel, Tha cost of complete first station of this type 48 approximately US$ 2500, including equipment for anelysis. — Addi- tonal stations within the sane city will cost only USS 1000 aince there Will be no need to duplicate the anelyticsl equipwent, Approximately “ US$ 1000 will be required each year for reagents, materials ané spare parts. Labour costs, transportation and the support of a good laboratory MILL be the main additional costs. The initial monitoring results obtained and the information collected should form the basis for expansion of the monitoring network to include the measurement of other pollutants. If possible, an approxinate dif- Fusion model should be made at this tine to guide the further development of the network, Examples of the development of air quality monitoring programmes are given in Annex 2. In addition to illustrating the person= hel and financial requirenents for different situations, these examples eaphasize the developaent of a programe in stages. 2.3 Design considerations ‘The types of preliminary information that are useful in the design of the monitoring programe were outlined above in section 2.1, Once this information is obtained and evaluated che baste framewrk of the monitoring programe can be developed, This will include: (G) selection of the pollutants to future; measured, initially and in the (41) definition of the area in vhich the measurements are to be wade; (ii) determination of the munber and distribution of the monitoring stations; and (iv) determination of the sampling duretion and frequency. ‘These decisions vill depend upon the objectives chat are to be ful- filled (See chapter 1) and the information to be provided. They vill also depend upon the phasing of the development of the programe, — In almost all instances the development of @ programe must be gradual, vith new pollutants being selected and the number of stations being reduced or increased over the years as the programe gains experience and is refined 0 respond more accurately to the need for which if war established. 2.3.1 Selection of pollutants ‘the selection of pollutants 4s comonly done in one of to ways. ‘the fret method recognizes chat the most comon air pollutants (see Annex 1) ave preseat 4a varying auounts in almost all urban areas. A ‘monitoring programme therefore usually begins with the measurement of Suspended particulate matter and sulfur dioxide; in many cases carbon monoxide 1s also mecsured if warranted by che traffie situation. The Programe ic then gradually developed to include other pollutants whose Presence might He suspected as 2 result of comparisons with other cities there they have already been confirmed by air quality meagurenenta, ‘This approach to pollutant selection is the most comon. It also indicates the need for standards where pollutant concentratfons are. show 9 be excesding or approaching levels considered to be haraful,20-23 or where levels still meet the criteria but a considerable increase in population or industrialization in the area Js anticipated. a9 ‘The second method for selection is less eapirical. It relies on emission faventories to identify not only the pollutante that could be potentially dangerous for human health and the enviroment in general but also the priorities if the programe is to be developed in stages. Ideally this approach is used in conjunction with modelling to caleulate the expected anbient Levels of various pollutants, Usually the firet eigaion aventories are done for the mejor pollutants and the programe iis then initiated. Subsequently inventories may be done for specific types of industries, which may be enitting particular pollutants that are fot common to all urban areas, for example, hydrogen sulfide, arsenic, Fluorine, asbestos, ete. Although, for purposes of clarification, two separate approaches have been ideacified for selecting air pollutance to be measured, in most cases the choice will be made seing a combination of the two. 2.3.2 Selection of the ares to be studied In most instances air sonitoring networks should encompass an entire urban complex. The reasons for this are obvious; air pollutants can travel long distances and their effects can occur many Kilometres from the point of enission, particularly in cases where the emissions are dis charged from high stacks, In order that the problen may be studied comprehensively, it is ox- treaely desirable that the entire air basin be included in the monitoring network, The area to be covered will, in aany cages, include more then one administrative region and this may require the development of co operative programmes with neighbouring communities. Certain characteristics of an urban region are important in the de~ finition of the study ares. Thus, for example, high mountain ranges oF large bodies of water can effectively serve ae boundaries of the area. In cases where a monitoring aysten should provide information about ‘ambient concentrations around a single source, the selection of the study area depends sainly on the stack height and the topographical snd meteoro- logical conditions, Some of the sempling sites should be loceted where the maximum aabient concentrations will be expected. In general, maximm ground level concentrations nay be expected dowavind from the source at a Aistance of 10-20 times the stack height 2.3.3 Humber and distribution of sampling aites ‘The number of sanpling sites depends on (1) the size of area to be covered by the network, (2) the variability of pollutant concentrations and (3) the data requitenents, which are related co the monitoring objectives. ‘Table 3 presents 2 general guide to the number of stations needed for monitoring trends of the common pollutants in urban areas, The popula tion figures are aseumed to be indicators of region size and pollution variability, The aunber of stations recomended is based largely on experience gained in some cities and should only be considered as a guide, Te includes a number of modifying factors chat rumsrize particular situations vhere the recomended suaber should be increased or decreased. For other monitoring objectives, particularly in relation to epidenio~ logical requirements, the auabers will usually have to be increased. 20 TABLE 3. SUGGESTED AVERAGE SIMSERS OF STATIONS FOR ATR QUALITY {TREND MONITORING IN URBAN AREAS OF GLVEN POPULATIONS” eee dioxide oxides Oxidants onoxide ane * woditying factors are as follows: (1) Im highly industrialized cities the aumber of stations for suspended particulate matter and sulfur dioxide should be increased. (2) tm areas where Large amounts of heavy fuel are used the number of ataclons for sulfur dioxide should be increased, (3) tm areas where not much heavy fuel is used the number of stations for sulfur dioxide may be reduced, (4) mm regior number of stations. with Lreegular terrain it may be necessary to increase the (5) Tm cities with extremely heavy traffic the nunber of stations for nitrogen oxides, oxidants and carbon monoxide may need to be doubled. (©) tm cities with a population of 4 million or more, with relatively low traffic, che number of statfons for nitrogen oxides, oxidants and carbon mmnoxide aay be reduced. The disteibution of @ given number of monitoring stations over s certain area can basically be of one of two types, (1) a geometrical net~ ‘work in which the sampling points are sited at the intersections of = grid or within each rectangle, or (2) a more solective network that refers te the choice of sites within the area to be surveyed, taking into consideration distribution of sources, population, et Goomserical netyprke have been used in a number of countries for = variety of purposes.222° In any cases auch networks were part of studies aiming to determine the number of stations required to charac~ Eerize adequately the air pollution level 1a an area; others are being ised co make routine air pollution measurements, ‘There are a nunber of possible variations regarding the operation of regular grid pattern network, ALL the cross pointe in the grid may have monitoring stations of ony some (sometimes because the others have been eliminated after data analysis has show that a reduction wae possible without loss in precision of data), There is also the possibility of using mobile sampling facilities to visit the grid points Following 4 (random) schedule using statistical procedures to calculate air pollution levels for the area. Although regular grid pattern networks axe being vied today, they’ are not generally favoured becuse of some Inherent shortcomings. They fare, for example, very Limited with regard to the saapling and analytical techniques that can be used economically. They alao tend to require & lot of staff time to operate, A more selective network involves the choice of sore representative sampling poines and generally affords a more detailed picture of air pollution levels in the most polluted and populated areas by increasing the monitoring sampling network density there, with @ concurrent reduction in the outskirts where the air quality is generally better, The number of sir quality sonitoring stations that can be installed will not general~ ly permit full resolution of ambient air pollution patterns for the region concerned, The use of dispersion modela, however, can provide additional infornation, through interpolation, for exanple, atout maxima pollution levels expected or the spatiel distribution of ambient concentrations be~ tween distant monitoring stations. Table 4 gives an example of che distribution of the sample sites be- tween the city centre oF industrial areas and residential areas. While single stations are often used to monitor several pollutants, ehis will not generally provide the greatest asount of information concerning the various types of pollutants, For example, aonitoring sites for sulfur Aioxide and suspended particulate matter can frequently be distributed fas shown in Table 4; however, a different grouping should be adopted for carbon monoxide monitors. Nearly all the latter should be placed in areas of high tratfic density, because carbon monoxide concentrations LIL be found to decrease rapidly avay from euch areas. On the other hand, the maximum concentrations of oxidants are Likely to be found at the downwind edge of a city, or beyond, because of the tine delay in- volved in the formation of these substances in the atmosphere, There Fore a monitoring programe for oxidants should slays include sone sites on the periphery of even outeide the urban region. TABLE 4. AN EXAMPLE OF THR DISTRIBUTION OF SAMPLING STATIONS Noaber of stations ‘otal aunber of cations palearicse in residential ee or industrial ee 1 1 ° 2 1 1 3 2 1 4 2 2 5 2 2 10 6 4 Mobile stations can often be used to augaent substantially the infor- mation available from fixed-eite networks, llowever, particular care must be exercised to sec that the equipment performance ir not changed during tthe changes in location. More frequent calibration will be required to guard against this. Also, the pollution distribution information from Such mobile sites is based on sanples usually collected over a small frac~ tion of the year. Therefore, precautions will have to de taken to ensure that the measurements represent actual pollution conditions during the periods of sampling, and that they are fot applied to other perfods of the year. 2.3.4 Sampling duration and frequency. ‘the duretion of a monitoring programe can be viewed on three tine seales: (1) short-term: (2) intemteeene; and (3) permanent. Short-tera monitoring is nomally carried out for special purposes auch ag a preliminery assessment of pollution levels prior to beginaing more extensive monitoring effort. | A limitation of this type of Tonitoring programme is that the data obtained way not be fully repee- Sentative because of the particular meteorological or emission conditions that may have prevailed during the sampling period, Intermittent monitoring (for example, | aonth per season or 1 day out of every 6) can ‘save costs, especially vhen manual sanplers are enployed. If the Sampling record 1s long enough (for example, several years) the data can be very useful for trend analysis and control strategy evaluation. However, care mist be exercised in not over-incerpreting data that may be biased due to am incomplete record, Permanent monitoring 1s usually carried out whenever an air pollution agency hat determined, on the basis Of either short-term or interaittent measurements, chat such sampling i= jeeded. In most cases, @ saall aonitoring network vill be established initially with a view Co increasing the nusber of stations in the future as further urbanization and induscrialtzation take place Te should be noted that with permanent monitoring long-tere trend analysis of the data is very important Co allov the agency to institute the necessary alr pollution control programme plenaing and legislation and/or regulations. Thus it fs essential chat the monitoring sites Teo main the sane for long periods of time. a With respect to sampling frequency, two factors play a doninant role, (2) the inherent variability of the pollucant (e.g., diurnal, dey of the week, seasonal), and (2) the required precision of the air quality data, ‘hich is related to the monitoring objective. Experience has show, for example, that concentrations of sulfur Atoxide and suspended particulate matter exhibit diurnal fluctuations that fare related to source eaiesion and typical dasly meteorological variations: Carvon monoxide concentrations, on the other hand, exhibit divrnal variations based primarily on traffic flov and density patterns. Si sonal differences in air pollution levels are related to source emission and meteorological variations. Weekday/weckend sampling has been shown to be very useful én assesoing industrial enissions as most industrial sources either curtail or cease operations during the weekend. Air pollution sampling during the weekend, Lf carried out long enough, can also give very useful information on the effects of automobile traffic on air pollution levels. ‘To determine characteristic pollutant concentration fluctuations sexpling should be more frequent than the expected frequency of variation, For exanple, co measure divrnal variation hourly samples should be taken. TE the measurements are not contiouous, the individual observations should be uniformly distributed throughout the day to get representative pattern. Similarly, to study = weekly pattern of daily means one should sple both on weekdays and at weekends. If an annual average is to be computed from the data, then st is essential that all portions of the year be represented equally. Recogni~ tion of the seasons] variation of measurenents made for certain pollutants shows why this is cesential, As a convenient rule, it may be assused that che nonitoring programe is adequately balanced if each calender quarter contains not’ less than 202 of the total number of observations nade within the year, ‘The accuracy of the average associated vith various sampling fre quencies has been calculated theoretically ggé determined enpirically at 2 tumber of locations throughout the world.® "Obviously the accuracy decreases with sampling frequency. For example, if samples are collected fovery other day, deviation from the annual average obtained by daily sampling is less chan + 2%, Sampling only every twelfth day yields a value within 4 10% of the average obtained with daily sanpling. If the air quality data are to be used for comparison with air quality standards or criteria the sampling poriod must be related to the averaging tine of the applicable standard.29 Tt must be stressed that Aitferene averaging times produce results that erenot directly comparable, The reason for this is the strong influence of the averaging time on the variance of a sample and through this on maximm and percentile Valuce.20 Wherever possible it is best to employ comparable sampling and averaging tines. En CHAPTER 3, STATION LocATTON CRETEREA ‘the exact location of each monitoring station is @ very inportant pect of the network development since a faulty choice may cause the data produced to be of very Limited value, possibly for years £0 come. ‘the location for an air quality monitoring station should satisfy the following: (Q)_ The site should be representative of the ares selected in the general design; (2) The station should be set up and operated s0 ax to yield data that can be conpared vith those from other stations within the network; (3) Gertain physical requirements should be satiefied at the site. ‘the ultimate choice for each site will be a compromise optiniaing these various considerations. 3.1 Representativenses A station is representative if the data obtained reflect concentration Levels and fluctuations of air pollutants within che given ares, in practice, the guidelines for this are difficult co specify. Whether the position of a station is satisfactory might, however, be checked by making simultaneous measurements at one oF more temporary St@~ tions within the ares concerned, The station should be located in @ place where interferences in the imediate vicinity are most unlikely, ies, away from: (1) nearby air pollution sources. the recomended distance will depend on the height and exission strength of the sources; the station should be at least 25m avay from domestic chinaeys, espectelly if the chimneys are lower than the sampling point; with larger sources the Gistance should be greater; (2) absorbing eurtaces (foltage and absorbing bullding materials), ‘me eleerance to be allowed will depend on the absorbing properties of the material for che pollutant in question, but ir will nomally be at least Lm; (3) areas where substantial rebuilding or land-use changes are fore- seen in the near future, especially if loag-tema trends are to be observed, ‘The guidelines above are valid for stations measuring general pollu tion levels and for most of che objectives Listed in Chepter 1, For objectives such as the study of health effects and the evalus- tion of danage to vegetation or building materiala, or the snvestigetion of specific complaints, the location of sampling stations is represen tative if the data obtained reflect the actual exposure of the receptor. Por hunan exposure this means that the sovements of @ population group ander study should be covered by a number of monitoring stations, in order to assecs the average exposure, There is a fundamental problem here, however, for air quality monitoring activities are generally conducted in outdoor air, wheres people normally spend auch time indsors. ‘The 2 concentrations of some pollutants, notably sulfur dioxide, are often Feduced through absorption on walls and other surfaces. On the other hhand, there ate circumstances, in which air pollution concentrations may be higher indoors than outdoors.?! The diversity of conditions affecting indoor concentrations 1a #0 great that it is ispracticable to recomend the tnclueion of indoor sampling as part of = general monitoring progranne. This feature mist, however, be considered in epidemtological studies ‘the measurenent of atx pollution from motor traffie needs special attention because of the sharp gradients in concentration Likely to be en countered. Ie may not need to de done at all of the network stations. As far ae primary pollutante from traffic are concerned (carbon monoxide, nitrogen oxides, hydrocarbons, Lead, and saoke from diesel vehicles), che highest concentrations are Likely to be found in busy strects in the centre of a city, particularly where there are tall bufldings along both Sides, Limiting the natural ventilation. If the street is mich frequented by pedestrians, then t may be appropriate to set up 2 sampling station neat by, if a suitable site can be found. A heterogeneous group of pollutants called photochemical oxidants, with ozone ae the sain component, occur in aabient air as a result of complex reactions between pollutants fron traffic, in certain atmospheric Conditions. As these requie tine to develop, concentrations of oxidants my be higher at or beyond the periphery of the urban area than within it. This should be taken into consideration when siting Instrumentation for the measurement of oxidant pollutants. 3.2 Requirements for comparability. ‘to f4cklitate the comparison of air quality dats obtained at dif~ ferent sites, the details of each location should be standardized as far a5 possible. Tf pollutants originating from stationary sources are being measured, the sampling intake should preferably be 3-4 m above che ground and 1-1.5a From the nearest vertical of horizontal surface, On al other sides it should be open, which means that the intake chould not be within a con~ Eined epace (closed courtyard of patio), in a corner, wader or above @ balcony, ete, For traffic pollution messurments the sanpling intake shovld be 3. above the strect level and at a horizontal distance of 1 a from the kerb. Deviations from these requirenents are critical and if they have to be made, they should be the same for all stations within a network. the height of 3m 4a recomended to avoid re-entraiment of particulates from the street, to permit free passage of pedestriens and to protect tho sampling intake from vandalism. Certain aonttoring instruments (e.g., high-volume samplers) have to be outdoors, and for practical reasons these are often located on the roofs of low buildings, caravans, ete. Instrumente for monitoring gaseous pollutants, as well as some of those for sampling suspended particulate matter (e.g. snoke measurement devices), are normally located indoors or in special shelters. They are connected with the outdoor air by s sampling tube ending with an inverted funnel, vbich serves to prevent the intake of precipitation and of large particles with the sample air, If suspended particulates are being Senpled the intake flow characteristics (flow rate and funnel élameter) 26 ‘must be standardized and the sampling Line should not have sharp bends. Te should also be ae short as possible, preferably no more than 3 a, and made of a material that does not react with the pollucant being sampled or Felease interfering vapours, The diameter of the tube will depend on the flow rate and ahould be standardized for the stole network. Since st ie often difficult to mect all the requtrenents Listed in sections 4,1 and 4.2, possible differences should be acknovledged and thelr Influence Tecognized, eepecially when making conparisons between the re- sults rom the various stations. Ie ie well know that ai pollution Levels in an urban area vary according to the general type of neighbourhood in which the measurement is made, For this reason it is useful to provide a description of the physical surrounding of each site that may be used uhen enalysing che Gata. Moat commonly, the description is made on a form especially de~ signed for this purpose, uhich, in addition to giving the address of the site, identifies whether it is located in the city centre or the suyurbs and vhether it is in an industrial, residential or comercial area. Additional information xegerding tie time zone, elevation above ground fand 20 on can aleo be given, 5.3 Bhysteal requirements ‘he aite where the station is located shovld folfil one or more of these requirements, depending on the type of inotruients used: (Q) te should be available for 2 long perio (2) te should preferably be accesstble 24 hours 2 day throughout the year; (3) Electrical power of suffietent rating should be available; (4) Te should be vandaiproofs (5) te may need to be protected from extrene tenperatures. Public buildings are often conventent and are frequently used for the siting of monitoring stations, a GHAPTER 4, OTEK ESSENTIAL COMPONENTS OF THE MONTTORING PROCRAOEE 4.1 Laboratory requirements ‘The operation of air quality monitoring networks will require the sup- port of laboratory facilities, The pollutants of interest, types of monitoring instruments used, size of network and frequency of sampling will Aictate the kind and extent of the Generally, the minimum equiment required for an air quality laboratory comprises: analytical belance, colorimeter, water distillation or ion exchange apparatus, air volume and air flowrate measurenent devices, laboratory glassware, drying oven, calibration equipment with appropriate standards, and, if possible, sufficient reagents and supplies for a year to ensure uniform quality. A reflectometer or transmissometer may be re- quired for the assessment of smoke stains on filter-papers, Spare sete BE sampling equipment should be available to ellow observations to be maintained in the event of breakdown Tn the initial stages of operation of a network, it 1s advisable to seck the advice and assistance of staff from # laboratory or international organization having experience in the relevant field. ven the analysis of the more comon pollutants my present difficulties for those not ex- Pericnced in this work, and che analysia of trace anounts of other pol- Jutanes requires elaborate and expensive facilities and highly skilled personsel, 4m all cases preference should be given co the use of internationally recomended standardized procedures, vith a view to achleving better coa- parability within the network and between networks ons nacional end international scale, 4.2 Quality assurance A qualicy assurance component is essential co any monitoring pro= grame in that it provides information sndicating whether the dats. pro- duced are acceptable according to procedures set out for this purpo: Since air pollution measurements are made by nusevous agencies ata large puaber of Sonitoring stations and laboratories, quality arsurance is also concerned uith establishing and assessing comparability of daca quality ‘anoag agencies. Qualigy sssurence may be divided into internal and external quality ‘aosurance. 3) Internal quality assurance may be understood to include oral internal procedures such as periodic calibrations, duplicate checks, Split samples, spiked samples and the keaping of adequate and neat records, External quality assurance aay be taken to include thove activities that are peforsed on a more occasional basis, usvally outside the orm] routine operations, e.g.) on-site system surveys, independent performance aedies, interlaboratory comparisons, and periodi¢ evaluation of internal quality assurance data, 4.3 Deva acquisition and handling procedurs For results from manually operated equipment it may he sufficient to collate the resuite on standard forms, but where continuous equipment ts Included in the network, costs versus benefits of an automatic data acquisition system should be considered, In many instances, however, pen recorders are satis{actory for data acquisition. 6 The flow of data should be handled in the following stages: (2) Assembly of raw data (strip charts, digital print-outs, Lists of senually written observations) and supporting data euch as flow ratea, volune aeter read-outs, observations of weather conditions, caltbration deta, renerks on operational particulars of instruments, change of reagents, dbservacions of exceptional events like agricultural burning, recalibra tion with or without adjustment of the appropriate instrument or repiace- sent of parte of the equipment, (@) Validation: rejection of erroneous daca, correction of data according to results of calibration and lov rates and proper coordination with time of observation, (3) Storage: filling in of standardized fora generation of computer print-outs and cap. cord punching and (4) Retrieval: develoment of systems to facilitate access to sets of data. (5) Analysis: application of statistical routines ained at pro ducing @ fev characteristic quantities, such as frequency distributions, seasonal variations, averages and standard devistions. Advanced data-handLing aystens are available that not only convert analogue signals co digital deta and atore then, but aleo permit automatic adjustment of zero and concentration range to be done and provide read-out of eritical instrument variables. Data may also be tranamitted by use of telephone Link from rence sites to a centeal computer for further pro- cessing and storage. Goncrally, it is advisable to make use of the experience of agencies already operating similar equipment prior to Scquiring a systen, and to stare vith @ single station to gain experience, 4-4 Data presentation Data presentation shovld pemnit interpretation of data in accordance with the initial objective of the monitoring network. Commonly they are presented as averages (arithuetic or geometric) and maximum values, a6 woll as frequency distributions, over a given period of time, Yearly averages will reveal long-tera trends, and aoathly means will reveal seasonal variations. Daily means will give greater resolution land will show day-of-week variations and the effect of changes in weather patterns, Hourly data ere useful to show varistions related to hunan activity and to the meteorological diurnal cycle. Cumulative frequency distributions of the data are useful, to show tho poreentage of tine a given vaive has been exceeded, and graphical prow sentations are very helpful in presenting Large volumes of data. TE sufficient data pointe are available they can be presented as Asoplaths. Concentration wind-roses are also useful: in combination with eopographical data they can provide information about the Location of pollution sources in the area, o ensure uniforatty of output across the world, a data reporting syeten weed by WHO is available for the handling of air quality cata.? Exanples of retrieval formats, such as a plot of monthly averages for & single polivtant and a cumilative frequency distefbueion of all values for 2 year, are also available. 2 LR 2 % 10. a. 12. BB 1. 15. 20 REVERENCES World Meteorological Organfzation, Urban climates, Geneva, 1970 (Gocunent No. 254,7-P,141) World Meteorological Organization, Observation and measurenent of ‘ataoegherie pollution, Geneva, 1974 (document No. 368) FiaLléa, C, C Global atmospheric monitoring. Environ. Sei. Technol. 9! 30-34 (1975) ASE quality in colected urban areas, 1973-1974, Geneva, World Hesith ‘Organization, 1976 (WHO Offset Publication No. 30) Selected nmechods of measuring air pollutants. Geneva, World Health Organization, 1975 (WHO OFfeat Publication No, 24) World Meteorological Organization. Operations manual for sampling ‘and analysis techniques for chenical constitvents in air and ‘peesipication, Geneva, 197% (document No. 299) Pen American Health Organization. Annex 1: Manual of operations. Ini Pan Auerican Air Pollution Monitoring Network. Report of zesults, June 1967-Desesber 1970, Lina, CEPIS, 1972, pp. 103-126 (cuts Technical Series No. 10) allay, W. et al. Environsental pollutants ~ selected analytical ‘acthods, london, buttervorths, 1975 (Scientific Committee on Probieas of the Environment (SCOPE) Report No. 6) Schaetder, 7., ed. Automatic air quality monitoring systems, Ansterdan, Bleevier, 197% International Conference on Sovironmental Sensing and Assessment, Las Vegas, 1975, Nev York, Institute of Electrical and Electronic Engineers, 1976, vol. 1 and 2 Munn, R. B. Global Environment Monitoring Systen (GEMS): Action plan ‘for phase 1, Toronto, International Council of Scientific Unions, 1973 (Setencific Comittee on Problens of the Knvironment (S00PE) Report No. 3) Suess, M. J. & Craxford, S. R., eds Manual on urban air quality manage- ‘Bent, Copeshagen, WHO Regional Office for Burope, 1976 International Atomic Energy Agency. Objectives and design of eavirontental monitoring propramies for radioactive conteainants, Vienna, 1975 (Safety Series No, 41) Environmental Instrumentation Group, Lawrence Herkeley Laboratory. Instrumentation for environmental monitoring, Berkeley, CA, University of California, 1975, vol. 1, IA and 2 Diwel, L. & 2Undoe#, 0, J. Behebung und Katastermussige Dolumentation der natesionen’luitfrender Scolfe in die Acaosphtee, Koln, Sanisdier narmchageteris Reales Te 16. uy 1. 20. a. 2. 23. 26, 25. 26. ar. 28. 29, 30. United States Environmental Protection Agency. Guide for compiling. ‘a comprehensive enission Javentory (revised), esearch Triangle Park, NC, 1973 (Publication No. APTD-1135) numer, D. B, Workbook of stnospheric dispersion cotimates, Cincinatti, OH, Public Health Service, 1970 (Public Health Service Publication No. 999-AP-26) World Meteorological Organization, Dispersion and forecasting of ‘siz pollution, Geneva, 1972 (WHO Technical Note No, 319) Delance, 0, M, An urbsn diffusion aodel for estimating long-term waives of afr quality, J. Alx gollut, Control Ass., 21: 16-19 as7) uo Technical Report Series, No. $06, 1972 (Report of a HHO Expert Comittee on Air Quality Criteria ané Guides for Urban Air Pollutants) World Health Organization, Health hazards of the human environment, Geneva, 1972 United States Enviromental Protection Agency. The woriéte ate ‘quality sanagesent standards, Washington, X¢, 1974, vols 1 and PP (epa Publications Wo. EPA=650/5-75-0018 and b) Tarses, R, 1, A mathenatical godel for relating air quality measure ‘aunts to air quality standards, Kesearch T¥iangle Park, NC, TBTI (office of Aix Programs Publication No, AP-89) Gorshko, 3, 2. [The optiqun number of sampling stations and the sanpling érequency for surveying urban air pollveicn)y ‘rudy Glavnoi Geofizicesko{ Observatorii, 254: 140-150 (171) Williane, J.D. & Ednisten, 8. G. An air resource management plan Zor the Nashville zetropolitan area, Cincinnati, OH, PuBLic Health Service, 1969 (Public Healeh Service Publication to. 999-AP-18) Steatmann, H, Basie principles of che management of gaseous alr pollutases, Staub-Reinhs Luft, 25(9): 10-16 (1965) Air pollution, Proceedings of the Sixth Intermatéonal Technical esting on Air Pollution Modeling and its Application, Satelite iovtiture, Frankfort a/M, 1975 (EXW0 Comittee on the Challenges to Modern Society publication N. 42) Noble, G. J. 6 Akland, G, G, Procedures for handling serometric daca, J, Air Pollut, Control ass., 23: 180-184 (1973) Pring, B. Anbient air quelity standards ond their application, In: Manual on urban air quality managenent, Copenhagen, WHO Regional Oéfice for Berope, 1976, pp. 31-86 Mecuive, J. & Holl, V. E. Relationship between concentrations of atsosgherie pollutants and averaging time, Atzospheric Environment, 5: 291-298 (1971) a1 31, Blersteker, K,, de Graaf, i, & Nase, A, G. Indoor pollution in Rotterdan hones, Int. J. Air Water Poliut., 9: 343-350 (aes) 52, ladvig, Fe I, Berg, Ne J. & Hofman, A. Jy The selection of sites for air pollutant gonitoring: paper io, 76-115 presented at eho G9EH Aunual Vecting of the Air Pollution Control ciation, Portland, OR, 1976 33, United States Environmental Protection Agency. Quality assurance handbook for air pollution measuresent eystene, Volume 1 = ‘principles, Research Triangle Park, NC, 1976 (HPA Publication Ko, #PA-600/5=76-005) 2 ANE 1 [PRINCIPAL SOURCES AND POLLUTANTS OF POSSIBLE CONCERN IN URRAN AREAS ‘Type of source Peel Main pollutants Domestic heating or Wood, peat, dung, Suspended particulate matter cooking ete. Carbon monoxide Oxides of nitrogen coat, Suspended particulate matter Sulfur dioxide Carbon monoxide Oxides of nitrogen Light off, gas Oxides of nétrogen Sulfur dionice Industrial boilers, Coal, heavy ofl Sulfur dioxide power plants Oxides of nitrogen Suspended particulate matter Manufacturing . Specific pollutants related to industrial processes nature of process, e.g. sulfur dioxide and sereaptans froa pulp mille, heavy metals from ‘suelters, Fluorides from aluninien smelters, iron oxide from steol vorks, Gust fon coment works, ete, ‘Transportation Cesoline Carbon monoxide Oxides of nitrogen ese Hydrocarbons oxidants” Diesel fuel Suspended particulate matter Oxides of nitrogen odours Sulfr dioxide * Secondary pollutant forsed in photochemical reactions ia the atmosphere involving other pollutants 3 amex 2 [EXAMPLES TO TLUUSTRATE THE GUIDELINES FOR THE DEVELOPMENT OF AN AIR POLLUTION MONTTORING PROGRAMME ‘his annex presents three examples to illustrate the guidclines for ‘the dovelopnent of an air pollution eonitoring programme given in the main part of the text, Two of these describe alternative approaches to the implementation of a monitoring programe for an urban area, one bat primarily on @ manual operation, with relatively low eapital costs, and the Other incorporating a fully autcaatic telenetered system, It mutt be Stressed that these examples are put forward as extremes of a wide range of possibilities. Automatic samplers could, for example, be used without the telemetry, which would decrease the capital cost but with a concomitant increase in personnel time to collect and evaluate the data, Altern tively, there could be one central eite with a complete range of manual and automated equipment, and more Limited observations at other sites. It is Intended that these descriptions of urban monitoring programes should Allustrate the need to develop a programme in stages, over several yesrs, fand to train local professional and support personnel £0 inplenent the programe, The third example provides an illustration of the steps for implemen tation of an air pollution monitoring programe in the vicinity of @ Planned or existing industrial source. It shows the Saportance of designing a programme thet is responsive to the details of topography and meteorology in the vicinity of each site, Specific suggestions for numbers of stations, bedgets and personnel reqirenents are presented in the examples, However, hese have been put forvard for illustrative purposes only. In practice, the details for any fair monitoring programe must be developed from the actual conditions found ia the area Co de monitored, Exmple 1, Starting an_air pollution monitoring programme based primarily on manually operated instruncate ‘The programe is for a city with slightly over two million inhabitants, xt has 12°000 industrial establishments, of which 2500 are potentially important emitters of air pollution. There is a pover plant buraing ofl and @ munteipel incinerator, Some garbage goes to a primitive sanitary Land#fil, where part of it burns spontancously. There ere about 15 000 molti-fantly buildings, with garbage incinerators, Mild climatic coidstions wean that heating is required only for brief periods. Traffic congestion is tevore én the central part of the city and many of the vehicles are poorly maintained. Meteorological data are avellable from two stations, one on the outskirts of the city and another at the nearby Objectives of the progrann (G2) To evaluate the general afr quality conditions in che efty and to provide che basis for a study of long-term trends of pollutant concentrations. 23 (2) To provide data for the subsequent development of an alr pollution prevention and control programe for the city. Implementation of the monitoring programme (QD) cottect and coltate background information relevant to the pollution problem, including data on the types, locations and emission Characteristics of stationary sources, the numbers and types of mobile sources, the meteorological data fron existing stations and the distri button OF complatats about air pollution, (2) Betablish contacts with other organizations, including the neteorological service and che university, that will be invited to parti- cipate ins joint effort with the health authority, (3) Install two stations for the measurenent of suspended pareiculete matter end aulfur dioxide, one at the city centre and the fother in an industrial zone, the precise sites being selected vith the help of the meteorological service, (4) Set up meteorological instruments at the two fix sampling (5) _ obtain a portable instrument for the messuranent of carbon monoxide to be operated on a specified schedule, in relation to pollution from eraftic, (6) Acquire a motor vehicle for servicing the fixed stations and carrying out the carbon monoxide survey (initially this might be on a shared basis with soe other service). (7) obtain laboretory equipaent for assessing the concentrations of suspended particulate matter and sulfur dioxide in air samples (@) conduct @ short baste course (tio weeks) for personnel of the ‘air pollution departaent and other professionals fos the meteorological service, university and industry involved in che monitoring programe; a specialist lecturer would be required for this. Second ye (1) Gontinee sessurenents started in the first year. (2) Prepare a simple diffusion model. (3) Provide travel funds to allow the director of the programe to visit monitoring networks in other cities. (4) Prey programe, Fe @ report on the First year of operation of the ‘third year (2) _ being the results of the diffusion sodel, select three additional sites to be equipped with monitors for suspended particulate matter end sulfur étoxide. 36 (2) consider the relocation of one or both meteorological instruments, in view of netvork expansion. (2) Assign one staff member to a well-established alr pollution agency for training for a period of three sonths (4) Peepare a report on the second year of operation and on the need for a control programme, Fourth year (2) Prepare a complete report at the end of three years of nessurmente analysing all the information collected and making reco mendations concerning: (a) future course of action for the programe; (b) further considerations on the need for a prevention snd control programe, and requisite legislation. Additso activities ‘As the monitoring programme is developed, tome control activities will be inplenented, especially to take care of complaints end to reduce heavy snoke extesions, Personnel ‘The following personnel will be assigned to the programe. An inapector is included in the staffing in the third and fourth years to pomit a control programe to he started concurrently with the monitoring programe. type of Munber of man-yeare personnel lat year 2nd year ded year th year Engineer 1 1 2 2 Chemise yr ye 1 1 Inspector 7 si 1 1 Assistant 1 1 2 2 Tectntesan 1 1 1 1 Driver af a 1 1 Total ‘ ‘ 8 8 3 Sumary of costs (1976 figures) ‘These costs do not include freight, duty and installation charges that nay be involved with equipment smported froa other countries and a safety factor of 2 may be required on the overall budget. Cost of personnel is not included due to large variations between Aifferent countries, This cost ean be calculated from the preceding table, vs § re Ist year 2nd year 3rd year th year Total Suspended particulate satter and S02 sampling 2 500 = 3 (000 eat eae 3) S00 equipment Portable CO monitor 2000002 . = 2000 Meteorological equipment 400000 = = - 6 000 Laboratory equipment 200000 = 1000 5003: 500 Reagents 500500 500 5002 00 Motor vehicle 40000 e = 6 000 Pct and naintenance 2000 3.000 3.000 3.000 11.000 ‘training 4000 3000 © 4000 = 11 000 Office furnteure and Ae he ae ies expenses Miscellaneous 1500 1300 1000 1.000 & 800 Total 24 500 8 800 13 500 6 00052 800 Hxauple 2. Development of an urban air pollution monitoring programme based prinaZily on altonated equipaont ‘the programe is for a city with the sane general characteristies as chat in Exanple 1, but with emissions and meteorological conditions such that there is a real risk of episodes of high pollution occurring. Tt is arsuned that # fully automated monitoring system capable of yielding real- Eime observations may be warranted, and that sufficient resources are available for this, Objectives of che programm (1) To assess the range of concentrations of pollutants resulting both from stationary sources and fron moor vehicles, in terms of average ‘and peak values, at selected points in the urban area. (2) to provide a basis for following long-term trends in pollution, (2) To guide the developnent of control strategies. (4) To develop and validate atmospheric dispersion models for application to all eniseions of pollution in the urban ares. (5) To activate emergency control procedures during episodes of high concentrations of poLlstants. Inplenentation of the nonitoring prograsme Hirst year (2) Review avatlable equipment for the network and prepare decailed specifications for measurement and computing equipment, together with plane for network operation. (2) Establish contact with ell pertinent agencies for the develop- nent of @ plan for the sonitoring network. (2) obtain sampling equipment for suspended particulate natcer (high-volume saspler, plus low-volume stoke sazpler) and sulfur dioxide for to manually operated stations; install and begin to operate before the end of the jeer. (4) obtain evo portable carbon monoxide sositoring instruments for ‘an exploratory survey of pollution from motor vehicles. (S) Develop data-reporting formats for 11 parameters. (6) Peepace fLeee approximation of sn eaisaion inventory for stationary sources and actor vehiel (7) determine detaited staffing requirements and develop plans for the eraining of necessary persoanel. (8) obtain two sete of meteorological equipnent, installing one on a atandsed 10 m tover close to one of the initial sampling sites and the other on a tall tower, e.g., 100, Lf @ suitable one exists. Second year (1) Parehase tuo sats of automatic continous instruments for measuring sulfur dioride, caides of nitrogen, oxidants and carbon monoxide, Enstalling these in the two stations already operating with manoal instruments; a0 computing equipment will Se purchased at thie time, the records on initial installation being maintained on pen recorders of data loggers. (2) Use the two Anktial stations to train cechnéeal and maintenance state, (2) taiesace a quality asterance and calsbration programe for the (4) expand the analysis laboratory to support the calibration 39 (5) Provide spectal training (the. Gtrector and the principal engineer. months) for the programme (6) carey out model calculations to assist in choosing locattone for the monitoring stations to be installed during the following year. Third year (QL) Purchase six additional sets of autoaatic aonitoring equipment (for aulfur dioxide, nitrogen oxides, oxidants and carbon monoxide), together vith computing equipment for the central analysis facility. Install at sites as indicated by the above calculations, but 4£ integrates stations are to be set up, vhere all pollutants are measured, then one should be at a site close to traffic, selected on the basis of the exploratory carbon monoxide survey, and another shovlé be on the leeward edge of the urban area, where relatively high concentrations of oxidants say be found, There may be no need to monitor carbon monoxide at ites avay from traffic, and it is assumed that this inetrument could be onicted ‘from a€ least tvo of tho sampling sites, (2) Purchase six addiefonal sets of equipment for measuring suspended particulate matter, and install at the new eites; these will require regular site visits by the technical staff, (3) Prepare conmuicstion Links between the sampling sites and the central analysis facility. (4) Conduct @ training programme for engineering sné technical stags Jigned to operate the network, Howreh year (2) _ Implement the operation of che communication Links and the computer in the central analysis facility; note that the operation of the two sete of manval instruments and on-site recording of outputs from automatic inetrumente are planaed to continue at least through the end of the fourth year, (2) Begin full operation of the network. (3) Inseseute regelar reporting procedar (4) Implement an episode control programe, using the real-time communication capability of the network, Additional activities Regular summaries of the observations will be prepared and the type of Legislation required and appropriate control techniques considered ‘during the third and fourth years, in the Light of resulte obtained from the manual and automatic instruments, Persoonel Personnel recomended for the systan include staff for data analysis and interpretation, as yell as staff for operation of the network, In the fourth year, whon the network will be in operation, @ total of 20 staff 1s recomended including tuo inspectors for control activities, 40 umber of man-years Type of personnel 16 year and year 3rd year Sth year Director 1 1 1 1 Instrument engineer 4 2 2 2 peta analyst % % 1 1h Computer operator - % 1 1 Computer programmer - 4 % 1 Analytical chemist 1 1 vi 2 Meteorologist 4 4 1 Inspector 5 = 2 2 ‘Technician 2 3 3 3 Secretary 1 2 2 3 Driver 1 1 1 1 ‘otal 7h 2 u 20% Sumary of costs (1976 figures) ‘The following external costs are estimated for the monitoring network, based on experience in the United States of Americ: Bizet year us § Manual sampling equipment for $0 and irtievlates, 2 ote 2 500 Portable oo sonitors, 2 4 500 Meteorological instruments, 2 sets, ‘with recorders and tower 30 000 Consultant astistance on systen design 10 000 Miscellaneous 3000 ‘otal 50 000 ond year Automatic monitoring instruments vith recorders for $02, 60, RO/M07 and oxidants, Deets; shelters for each set included 100 000 Laboratory instruments and supplies 40 000 ‘raining for dixector and principal engineer 10 000 Motor vehicles with expenses 14 000 office supplies 3 000 Miscellaneovs 3 000 ‘otal 170 000 a third year us § Autanatie monitoring ingtruaents with shelters, 6 sets 280 000 ‘small computer 20 00 Communication interface equipnent for 8 neations 24 000 short course for staff 15 000 Vehicle eintenance 6 000 Total 345 000 Fourth year Vehicle maintenance 6 000 Office eupplies 2 000 Conminseation Line cost to connect 8 stations co central facility 16 000 Total 24 000 otal extemal coste, four yeare 589 000 Contingency (10%) for other equipment and supplies 59 000 ‘otal 648 000 Because of uncertaintios in the costs of importing and installing ‘equipnent in various countries, a safety factor of 2 1s recomended for this cout estimate. Therefore, the external costs for the eysten are estimated to be in the range of USS 650 000-1 300 000. In addition, costs for personnel and facilities must be included in the total budj Provision may algo need to be made for the purchase of at least one spare set of equipment to maintain observations in the event of failure Example 3, Developing an sir quality monitoring programe in the vicinity of a iatge single source the large single source {2 a 500-tH conl-fired pover station situated along a river on the outskirts of a medium-sized city. Most of the other industrial activity in the urban area consists of secondary production plants with relatively low air pollution emission rates. Consideration is currently being given to doubling the capacity of the power station. objectives of the programe (2) To determine the distribution of air pollution emanating from the power station in time and space. (2) To provide a data base for calculating the impact of the expansion of the station on current anbient air quality levels, 42 inp enentation of the sonitoring programe Firat stage (2) kstimate the enlesion rater of the various air pollutente enitted fron the pover station stacks, notably sulfur dioxide and particulates. (2) Study the topographical map of the region, noting ridges and valleys that may affect the dispersion of the air pollutants, (3) Undertake @ site inspection, noting land use and vegetation patteras and possible signs of vegetation damage, and classify complaints by area, (4) Prepare a winé-rose for the axes, based on Local meteorological observations; an experienced meteorologist should review the repre Sentativeness of chis wind-rose in locations having uneven terrsin, If sufficient meteorological observations are available, conpute the pollution potential for the area, taking special note of the frequency of Linited ‘mixing conditions (i.e., inversions based just above the expected plune height). (5) Undertake diffusion model calculations for various meteoralo- ‘sical conditions to obtain estimates of ground level concentrations preferably in the forn of frequency distributions of hourly values at fixed grid points surrounding the poxer plant. Second stage (1) 0m the basis of the diffusion calculations and any evidence of vegetation damage, select a network of stations and pollutente to be monitored, For major single-source aonitoring @ network of three co six Stations Ls frequently (but not alvmye) chosen. Fever than thes stations would be inadequate for the task, and date from nore than six stations can have Limited use, except in large-scale programmes, Concentrations of ambient sulfur dioxide and svepended particulate matter are uormally ‘monitored in assessment programes for coal-fired power plants. Nitrogen oxides and trace metals in the airborne particulates are sizo measured in sone programmes. (2) install one or more continuously-recording aneaometers at the appropriate sites, One anenoneter should be at standard exposure height (Gon) in flat, open surroundings, Eo permit comparisons with other wind observation records. A second anqnometer 1s often installed on a tower at an elevation similar to the stack top; this permits evaluation of representative wind dispersion conditions for the elevated plane, (2)__Arrange for the estination of emission dats f:om the pover station stack, This can be provided by direct meamurenent in the stack Gr estinated from data on fuel use and/or boiler operation, (4) Use the monitoring data and the entesion estimates to calibrate the dispersion model, Cases of high concentrations should be specially evaluated to describe the conditions thet produce maximum impact, (5) Forecast the inpact of the pover station expansion on anbéent air quality levels in the region. 3 Personnel The conduct of the programe dercribed sbove will require the servic: of an engineer, a meteorologist, a computer programmer, one or tvo technicians and a secretary, fhe total time required for both stages WiLl frequently be four to six months, Hovever, it may be desirable to spread the actual monitoring activity over an entire year (e.g-, 1 month of field operation in each of the four seasons) in order to evaluate the changes in impact throughout an annval cycle. Gosee Ansuning that one or two anenoneter sites and fron three to six monitoring sites for sulfur dioxide and suspended particulate aetter will be established, the equipment costs will be about US$ 35 000-70 000. Conputer services and expendable supplics for the instruments will add $8. 5000-15 000 to the basic cost, Personnel costs are not included in these estinates. ane 3 LIST OF PARTICIPANTS 1 THE WHO/WMO CONSULTATION ON AIR QUALITY MONITORING IN URBAN AND INDUSTRIAL AREAS, GENEVA, 21-25 JUNE 1976 Mr G, Akland, United States Enviromental Protectfon Agency, Research ‘TeLangle Park, NC, USA Dr K, Blersteker, Municipal Heaith Departnent, Rotterdam, Netherlands (riee-chaimnes} Mr F, Bormann, United Scates Environnental Protection Agency, Research ‘Telangle Park, NC, USA Mo M, Fugas, Institute for Medical Research and Oceupational Health, Zagred, Yugoslavia Dr O.-H. Killingno, Aix Protection Laboratory, Nyktping, Sweden Dr A, Kolnala, Finnish Meteorological Institue, Helsinki, Finland Dr J. Mahoney, BRT International, Concord, MA, USA Dr Ry Yun, Department of the Environment, Domsview, Ontario, Canada

You might also like