Air quality monitoring means the systematic measurement of ambient air pollutants in order to be able to assess the exposure of vulnerable receptors (e.g., people, animals, plants and art works) on the basis of standards and guidelines derived from observed effects, and/or to establish the source of the air pollution (causal analysis).

Ambient air pollutant concentrations are influenced by the spatial or time variance of emissions of hazardous substances and the dynamics of their dispersion in the air. As a consequence, marked daily and annual variations of concentrations occur. It is practically impossible to determine in a unified way all these different variations of air quality (in statistical language, the population of air quality states). Thus, ambient air pollutant concentrations measurements always have the character of random spatial or time samples.

Measurement Planning

The first step in measurement planning is to formulate the purpose of the measurement as precisely as possible. Important questions and fields of operation for air quality monitoring include:

Area measurement:

• representative determination of exposure in one area (general air monitoring)
• representative measurement of pre-existing pollution in the area of a planned facility (permit, TA Luft (Technical instruction, air))
• smog warning (winter smog, high ozone concentrations)
• measurements in hot spots of air pollution to estimate maximum exposure of receptors (EU-NO₂ guideline, measurements in street canyons, in accordance with the German Federal Immission Control Act)
• checking the results of pollution abatement measures and trends over time
• screening measurements
• scientific investigations - for example, the transport of air pollution, chemical conversions, calibrating dispersion calculations.

Facility measurement:

• measurements in response to complaints
• ascertaining sources of emissions, causal analysis
• measurements in cases of fires and accidental releases
• checking success of reduction measures
• monitoring factory fugitive emissions.

The goal of measurement planning is to use adequate measurement and assessment procedures to answer specific questions with sufficient certainty and at minimum possible expense.

An example of the parameters that should be used for measurement planning is presented in table 1, in relation to an assessment of air pollution in the area of a planned industrial facility. Recognizing that formal requirements vary by jurisdiction, it should be noted that specific reference here is made to German licensing procedures for industrial facilities.

Table 1. Parameters for measurement planning in measuring ambient air pollution concentrations (with example of application)

The example in table 1 shows the case of a measurement network that is supposed to monitor the air quality in a specific area as representatively as possible, to compare with designated air quality limits. The idea behind this approach is that a random choice of measurement sites is made in order to cover equally locations in an area with varying air quality (e.g., living areas, streets, industrial zones, parks, city centres, suburbs). This approach may be very costly in large areas due to the number of measurement sites necessary.

Another conception for a measurement network therefore starts with measurement sites that are representatively selected. If measurements of differing air quality are conducted in the most important locations, and the length of time that the protected objects remain in these “microenvironments” is known, then the exposure can be determined. This approach can be extended to other microenvironments (e.g., interior rooms, cars) in order to estimate the total exposure. Diffusion modelling or screening measurements can help in choosing the right measurement sites.

A third approach is to measure at the points of presumed highest exposure (e.g., for NO₂ and benzene in street canyons). If assessment standards are met at this site, there is sufficient probability that this will also be the case for all other sites. This approach, by focusing on critical points, requires relatively few measurement sites, but these must be chosen with particular care. This particular method risks overestimating real exposure.

The parameters of measurement time period, assessment of the measurement data and measurement frequency are essentially given in the definition of the assessment standards (limits) and the desired level of certainty of the results. Threshold limits and the peripheral conditions to be considered in measurement planning are related. By using continuous measurement procedures, a resolution that is temporally almost seamless can be achieved. But this is necessary only in monitoring peak values and/or for smog warnings; for monitoring annual mean values, for example, discontinuous measurements are adequate.

The following section is dedicated to describing the capabilities of measurement procedures and quality control as a further parameter important to measurement planning.

Quality Assurance

Measurements of ambient air pollutant concentrations can be costly to conduct, and results can affect significant decisions with serious economic or ecological implications. Therefore, quality assurance measures are an integral part of the measurement process. Two areas should be distinguished here.

Procedure-oriented measures

Every complete measurement procedure consists of several steps: sampling, sample preparation and clean-up; separation, detection (final analytical step); and data collection and assessment. In some cases, especially with continuous measurement of inorganic gases, some steps of the procedure can be left out (e.g., separation). Comprehensive adherence to procedures should be strived for in conducting measurements. Procedures that are standardized and thus comprehensively documented should be followed, in the form of DIN/ISO standards, CEN standards or VDI guidelines.

User-oriented measures

Using standardized and proven equipment and procedures for ambient air pollutant concentration measurement cannot alone ensure acceptable quality if the user does not employ adequate methods of quality control. The standards series DIN/EN/ISO 9000 (Quality Management and Quality Assurance Standards), EN 45000 (which defines the requirements for testing laboratories) and ISO Guide 25 (General Requirements for the Competence of Calibration and Testing Laboratories) are important for user-oriented measures to ensure quality.

Important aspects of user quality control measures include:

• acceptance and practice of the content of the measures in the sense of good laboratory practice (GLP)
• correct maintenance of measurement equipment, qualified measures to eliminate disruptions and ensure repairs
• carrying out calibrations and regular checking to ensure proper functioning
• carrying out interlaboratory testing.

Measurement Procedures

Measurement procedures for inorganic gases

A wealth of measurement procedures exists for the broad range of inorganic gases. We will differentiate between manual and automatic methods.

Manual procedures

In the case of manual measurement procedures for inorganic gases, the substance to be measured is normally adsorbed during the sampling in a solution or solid material. In most cases a photometric determination is made after an appropriate colour reaction. Several manual measurement procedures have special significance as reference procedures. Because of the relatively high personnel cost, these manual procedures are conducted only rarely for field measurements today, when alternative automatic procedures are available. The most important procedures are briefly sketched in table 2.

Table 2. Manual measurement procedures for inorganic gases

DL = detection limit; s = standard deviation; rel. s = relative s.

A special sampling variant, used primarily in connection with manual measurement procedures, is the diffusion separation tube (denuder). The denuder technique is aimed at separating the gas and particle phases by using their different diffusion rates. Thus, it is often used on difficult separation problems (e.g., ammonia and ammonium compounds; nitrogen oxides, nitric acid and nitrates; sulphur oxides, sulphuric acid and sulphates or hydrogen halides/halides). In the classic denuder technique, the test air is sucked through a glass tube with a special coating, depending on the material(s) to be collected. The denuder technique has been further developed in many variations and also partially automated. It has greatly expanded the possibilities of differentiated sampling, but, depending on the variant, it can be very laborious, and proper utilization requires a great deal of experience.

Automated procedures

There are numerous different continuous measuring monitors on the market for sulphur dioxide, nitrogen oxides, carbon monoxide and ozone. For the most part they are used particularly in measurement networks. The most important features of the individual methods are collected in table 3.

Table 3. Automated measurement procedures for inorganic gases

It should be emphasized here that all automatic measurement procedures based on chemical-physical principles must be calibrated using (manual) reference procedures. Since automatic equipment in measurement networks often runs for extended periods of time (e.g., several weeks) without direct human supervision, it is indispensable that their correct functioning is regularly and automatically checked. This generally is done using zero and test gases that can be produced by several methods (preparation of ambient air; pressurized gas cylinders; permeation; diffusion; static and dynamic dilution).

Measurement procedures for dust-forming air pollutants and its composition

Among particulate air pollutants, dustfall and suspended particulate matter (SPM) are differentiated. Dustfall consists of larger particles, which sink to the ground because of their size and thickness. SPM includes the particle fraction that is dispersed in the atmosphere in a quasi-stable and quasi-homogenous manner and therefore remains suspended for a certain time.

Measurement of suspended particulate matter and metallic compounds in SPM

As is the case with measurements of gaseous air pollutants, continuous and discontinuous measurement procedures for SPM can be differentiated. As a rule, SPM is first separated on glass fibre or membrane filters. It follows a gravimetric or radiometric determination. Depending on the sampling, a distinction can be made between a procedure to measure the total SPM without fractionation according to the size of the particles and a fractionation procedure to measure the fine dust.

The advantages and disadvantages of fractionated suspended dust measurements are disputed internationally. In Germany, for example, all threshold limits and assessment standards are based on total suspended particulates. This means that, for the most part, only total SPM measurements are performed. In the United States, on the contrary, the so-called PM-10 procedure (particulate matter £ 10μm) is very common. In this procedure, only particles with an aerodynamic diameter up to 10 μm are included (50 per cent inclusion portion), which are inhalable and can enter the lungs. The plan is to introduce the PM-10 procedure into the European Union as a reference procedure. The cost for fractionated SPM measurements is considerably higher than for measuring total suspended dust, because the measuring devices must be fitted with special, expensively constructed sampling heads that require costly maintenance. Table 4 contains details on the most important SPM measurement procedures.

Table 4. Measurement procedures for suspended particulate matter (SPM)

Recently, automatic filter changers have also been developed that hold a larger number of filters and supply them to the sampler, one after another, at timed intervals. The exposed filters are stored in a magazine. The detection limits for filter procedures lie between 5 and 10 μg/m³ of dust, as a rule.

Finally, the black smoke procedure for SPM measurements has to be mentioned. Coming from Britain, it has been incorporated into EU guidelines for SO₂ and suspended dust. In this procedure, the blackening of the coated filter is measured with a reflex photometer after the sampling. The black smoke values that are thus photometrically obtained are converted into gravimetric units (μg/m³) with the help of a calibration curve. Since this calibration function depends to a high degree on the composition of the dust, especially its soot content, the conversion into gravimetric units is problematic.

Today, metal compounds are often routinely determined in suspended dust immission samples. In general, the collection of the suspended dust on filters is followed by a chemical dissolution of the separated dusts, since the most common final analytical steps presuppose converting the metallic and metalloid compounds in an aqueous solution. In practice, the most important methods by far are atom absorption spectroscopy (AAS) and spectroscopy with plasma excitation (ICP-OES). Other procedures for determining metallic compounds in suspended dust are x-ray fluorescence analysis, polarography and neutron activation analysis. Although metallic compounds have been measured for more than a decade now as a component of SPM in outside air at certain measurement sites, important unanswered questions remain. Thus the conventional sampling by separating the suspended dust on filters assumes that the separation of the heavy metal compounds on the filter is complete. However, earlier indications have been found in the literature questioning this. The results are very heterogeneous.

A further problem lies in the fact that different compound forms, or single compounds of the respective elements, cannot be distinguished in the analysis of metallic compounds in suspended dust using the conventional measurement procedures. While in many cases adequate total determinations can be made, a more thorough differentiation would be desirable with certain especially carcinogenic metals (As, Cd, Cr, Ni, Co, Be). There are often big differences in the carcinogenic effects of elements and their individual compounds (e.g., chromium compounds in oxidation levels III and VI - only those in level VI are carcinogenic). In such cases a specific measurement of the individual compounds (species analysis) would be desirable. Despite the significance of this problem, only first attempts at species analysis are being made in measurement technique.

Measurement of dustfall and metallic compounds in dustfall

Two fundamentally different methods are used to collect dustfall:

• sampling in collecting vessels

• sampling on adhesive surfaces.

A popular procedure for measuring dustfall (deposited dust) is the so-called Bergerhoff procedure. In this procedure the entire atmospheric precipitation (dry and wet depositions) is collected over 30± 2 days in vessels about 1.5 to 2.0 metres above the ground (bulk deposition). Then the collecting vessels are taken to the lab and prepared (filtered, water evaporated, dried, weighed). The result is calculated on the basis of the surface area of the collecting vessel and exposure time in grams per square meter and day (g/m²d). The relative detection limit is 0.035 g/m²d.

Additional procedures for collecting dustfall include the Liesegang-Löbner device and methods which collect the deposited dust on adhesive foils.

All measurement results for dustfall are relative values that depend on the apparatus used, as the dust separation is influenced by the flow conditions at the device and other parameters. The differences in the measurement values obtained with the different procedures can reach 50 per cent.

Also important is the composition of the deposited dust, such as the content of lead, cadmium and other metallic compounds. The analytical procedures used for this are basically the same as those used for suspended dust.

Measuring special materials in dust form

Special materials in dust form include asbestos and soot. Collecting fibres as air pollutants is important since asbestos has been classified as a confirmed carcinogenic material. Fibres with a diameter of D ≤ 3μm and a length of L ≥ 5μm, where L:D ≥ 3, are considered carcinogenic. Measurement procedures for fibrous materials consist of counting, under the microscope, fibres that have been separated on filters. Only electron microscopic procedures can be considered for outside air measurements. The fibres are separated on gold-coated porous filters. Prior to assessment in an electron scan microscope, the sample is freed of organic substances through plasma incineration right on the filter. The fibres are counted on part of the filter surface, randomly chosen and classified by geometry and type of fibre. With the help of energy dispersive x-ray analysis (EDXA), asbestos fibres, calcium sulphate fibres and other inorganic fibres can be differentiated on the basis of elemental composition. The entire procedure is extremely expensive and requires the greatest care to achieve reliable results.

Soot in the form of particles emitted by diesel motors has become relevant since diesel soot was also classified as carcinogenic. Because of its changing and complex composition and because of the fact that various constituents are also emitted from other sources, there is no measurement procedure specific to diesel soot. Nevertheless, in order to say something concrete about the concentrations in ambient air, soot is conventionally defined as elemental carbon, as a part of total carbon. It is measured after sampling and an extraction step and/or thermal desorption. Determination of the carbon content ensues through burning in an oxygen stream and coulometric titration or non-dispersive IR detection of the carbon dioxide formed in the process.

The so-called aethalometer and the photoelectric aerosol sensor are also used for measuring soot, in principle.

Measuring Wet Depositions

Together with dry deposition, wet deposition in rain, snow, fog and dew constitute the most important means by which harmful materials enter the ground, water or plant surfaces from the air.

In order to clearly distinguish the wet deposition in rain and snow (fog and dew present special problems) from the measurement of total deposition (bulk deposition, see section “Measurement of dustfall and metallic compounds” above) and dry deposition, rain catchers, whose collection opening is covered when there is no rain (wet-only sampler), are used for sampling. With rain sensors, which mostly work on the principle of conductivity changes, the cover is opened when it starts to rain and closed again when the rain stops.

The samples are transferred through a funnel (open area approx. 500 cm² and more) into a darkened and if possible insulated collection container (of glass or polyethylene for inorganic components only).

In general, analysing the collected water for inorganic components can be done without sample preparation. The water should be centrifuged or filtered if it is visibly cloudy. The conductivity, pH value and important anions (NO₃ ^– , SO₄ ^2– , Cl^–) and cations (Ca²⁺, K⁺, Mg²⁺, Na⁺, NH₄ ⁺ and so on) are routinely measured. Unstable trace compounds and intermediate states like H₂O₂ or HSO₃ ^– are also measured for research purposes.

For analysis, procedures are used that are generally available for aqueous solutions such as conductometry for conductivity, electrodes for pH values, atom adsorption spectroscopy for cations (see section “Measuring special materials in dust form”, above) and, increasingly, ion exchange chromatography with conductivity detection for anions.

Organic compounds are extracted from rain water with, for example, dichloromethane, or blown out with argon and adsorbed with Tenax tubes (only highly volatile materials). The materials are then subjected to a gas chromatographic analysis (see “Measurement procedures for organic air pollutants”, below).

Dry deposition correlates directly with ambient air concentrations. The concentration differences of airborne harmful materials in rain, however, are relatively small, so that for measuring wet deposition, wide-mesh measuring networks are adequate. Examples include the European EMEP measurement network, in which the entry of sulphate and nitrate ions, certain cations and precipitation pH values are collected in approximately 90 stations. There are also extensive measurement networks in North America.

Optical Long-Distance Measurement Procedures

Whereas the procedures described up to now catch air pollution at one point, optical long-distance measuring procedures measure in an integrated manner over light paths of several kilometres or they determine the spatial distribution. They use the absorption characteristics of gases in the atmosphere in the UV, visible or IR spectral range and are based on the Lambert-Beer law, according to which the product of light path and concentration are proportional to the measured extinction. If the sender and receiver of the measuring installation change the wavelength, several components can be measured in parallel or sequentially with one device.

In practice, the measurement systems identified in table 5 play the biggest role.

Table 5. Long-distance measurement procedures

LIDAR = Light detection and ranging; DIAL = differential absorption LIDAR.

Measurement Procedures for Organic Air Pollutants

The measurement of air pollution containing organic components is complicated primarily by the range of materials in this class of compounds. Several hundred individual components with very different toxicological, chemical and physical characteristics are covered under the general title “organic air pollutants” in the emissions registers and air quality plans of congested areas.

Especially due to the great differences in potential impact, collecting relevant individual components has more and more taken the place of previously used summation procedures (e.g., Flame Ionization Detector, total carbon procedure), the results of which cannot be assessed toxicologically. The FID method, however, has retained a certain significance in connection with a short separation column to separate out methane, which is photochemically not very reactive, and for collecting the precursor volatile organic compounds (VOC) for the formation of photo-oxidants.

The frequent necessity of separating the complex mixtures of the organic compounds into relevant individual components makes measuring it virtually an exercise in applied chromatography. Chromatographic procedures are the methods of choice when the organic compounds are sufficiently stable, thermally and chemically. For organic materials with reactive functional groups, separate procedures that use the functional groups’ physical characteristics or chemical reactions for detection continue to hold their ground.

Examples include using amines to convert aldehydes to hydrazones, with subsequent photometric measurement; derivatization with 2,4-dinitrophenylhydrazine and separation of the 2,4-hydrazone that is formed; or forming azo-dyes with p-nitroaniline for detecting phenols and cresols.

Among chromatographic procedures, gas chromatography (GC) and high-pressure liquid chromatography (HPLC) are most frequently employed for separating the often complex mixtures. For gas chromatography, separation columns with very narrow diameters (approx. 0.2 to 0.3 mm, and approx. 30 to 100 m long), so-called high-resolution capillary columns (HRGC), are almost exclusively utilized today. A series of detectors are available for finding the individual components after the separation column, such as the above-mentioned FID, the ECD (electron capture detector, specifically for electrophilic substitutes such as halogen), the PID (photo-ionization detector, which is especially sensitive to aromatic hydrocarbons and other p-electron systems), and the NPD (thermo-ionic detector specifically for nitrogen and phosphorus compounds). The HPLC uses special through-flow detectors which, for example, are designed as the through-flow cuvette of a UV spectrometer.

Especially effective, but also especially expensive, is the use of a mass spectrometer as a detector. Really certain identification, especially with unknown mixtures of compounds, is often possible only through the mass spectrum of the organic compound. The qualitative information of the so-called retention time (time the material remains in the column) that is contained in the chromatogram with conventional detectors is supplemented with the specific detection of the individual components by mass fragmentograms with high detection sensitivity.

Sampling must be considered before the actual analysis. The choice of sampling method is determined primarily by volatility, but also by expected concentration range, polarity and chemical stability. Furthermore, with non-volatile compounds, a choice must be made between concentration and deposition measurements.

Table 6 provides an overview of common procedures in air monitoring for active enrichment and chromatographic analysis of organic compounds, with examples of applications.

Table 6. Overview of common chromatographic air quality measurement procedures of organic compounds (with examples of applications)

GC = gas chromatography; GCMS = GC/mass spectroscopy; FID = flame ionization detector; HRGC/ECD = high resolution GC/ECD; ECD = electron capture detector; HPLC = high performance liquid chromatography. PID = photo-ionization detector.

Deposition measurements of organic compounds with low volatility (e.g., dibenzodioxins and dibenzofurans (PCDD/PCDF), polycyclic aromatic hydrocarbons (PAH)) are gaining in importance from the perspective of environmental impact. Since food is the main source of human intake, airborne material transferred onto food plants is of great significance. There is, however, evidence that material transfer by way of particulate deposition is less important than dry deposition of quasi-gaseous compounds.

For measuring total deposition, standardized devices for dust precipitation are used (e.g., Bergerhoff procedure), which have been slightly modified by darkening as a protection against the entry of strong light. Important technical measurement problems, such as the resuspension of already separated particles, evaporation or possible photolytic decomposition, are now being systematically researched in order to improve the less-than-optimal sampling procedures for organic compounds.

Olfactometric Investigations

Olfactometric immission investigations are used in monitoring to quantify odour complaints and to determine baseline pollution in licensing procedures. They serve primarily to assess whether existing or anticipated odours should be classified as significant.

In principle, three methodological approaches can be differentiated:

• measurement of the emission concentration (number of odour units) with an olfactometer and subsequent dispersion modelling

• measurement of individual components (e.g., NH₃) or mixtures of compounds (e.g., gas chromatography of gases from landfills), if these adequately characterize the odour

• odour determinations by means of inspections.

The first possibility combines emission measurement with modelling and, strictly speaking, cannot be classified under the term air quality monitoring. In the third method, the human nose is used as the detector with significantly reduced precision as compared to physical-chemical methods.

Details of inspections, measurement plans and assessing the results are contained, for example, in the environmental protection regulations of some German states.

Screening Measurement Procedures

Simplified measurement procedures are sometimes used for preparatory studies (screening). Examples include passive samplers, test tubes and biological procedures. With passive (diffusive) samplers, the material to be tested is collected with freely flowing processes such as diffusion, permeation or adsorption in simple forms of collectors (tubes, plaques) and enriched in impregnated filters, meshes or other adsorption media. So-called active sampling (sucking the sample air through a pump) thus does not occur. The enriched quantity of material, analytically determined according to definite exposure time, is converted into concentration units on the basis of physical laws (e.g., of diffusion) with the help of collection time and the collector’s geometric parameters. The methodology stems from the field of occupational health (personal sampling) and indoor air measurement, but it is increasingly being used for ambient air pollutant concentration measurements. An overview can be found in Brown 1993.

Detector tubes are often used for sampling and quick preparatory analysis of gases. A certain test air volume is sucked through a glass tube that is filled with an adsorptive reagent that corresponds with the test objective. The contents of the tube change colour depending on the concentration of the material to be determined that is present in the test air. Small testing tubes are often used in the field of workplace monitoring or as a quick procedure in cases of accidents, such as fires. They are not used for routine ambient air pollutant concentration measurements due to the generally too high detection limits and too limited selectivity. Detector testing tubes are available for numerous materials in various concentration ranges.

Among the biological procedures, two methods have become accepted in routine monitoring. With the standardized lichen exposure procedure, the mortality rate of the lichen is determined over the exposure time of 300 days. In another procedure, French pasture grass is exposed for 14±1 days. Then the amount of growth is determined. Both procedures serve as summary determinations of air pollutant concentration effects.

Air Quality Monitoring Networks

Around the world, the most varied types of air quality networks are utilized. A distinction should be drawn between measurement networks, consisting of automatic, computer-controlled measuring stations (measurement containers), and virtual measurement networks, which only define the measurement locations for various types of air pollutant concentration measurements in the form of a preset grid. Tasks and conceptions of measurement networks were discussed above.

Continuous monitoring networks

Continuously operating measurement networks are based on automatic measuring stations, and serve primarily for air quality monitoring of urban areas. Measured are air pollutants such as sulphur dioxide (SO₂), dust, nitrogen monoxide (NO), nitrogen dioxide (NO₂), carbon monoxide (CO), ozone (O₃), and to an extent also the sum of the hydrocarbons (free methane, C_nH_m) or individual organic components (e.g., benzene, toluene, xylenes). In addition, depending on need, meteorological parameters such as wind direction, wind speed, air temperature, relative humidity, precipitation, global radiation or radiation balance are included.

The measuring equipment operated in measurement stations generally consists of an analyser, a calibration unit, and control and steering electronics, which monitors the whole measuring equipment and contains a standardized interface for data collection. In addition to the measurement values, the measuring equipment supplies so-called status signals on errors and the operating status. The calibration of the devices is automatically checked by computer at regular intervals.

As a rule, the measurement stations are connected with fixed data lines, dial connections or other data transfer systems to a computer (process computer, workstation or PC, depending on the scope of the system) in which the measurement results are entered, processed and displayed. The measurement network computers and, if necessary, specially trained personnel monitor continuously whether various threshold limits are exceeded. In this manner critical air quality situations can be recognized at any time. This is very important, especially for monitoring critical smog situations in winter and summer (photo-oxidants) and for current public information.

Measurement networks for random sample measurements

Beyond the telemetric measurement network, other measuring systems for monitoring air quality are used to varying extents. Examples include (occasionally partially automated) measurement networks to determine:

• dust deposition and its components
• suspended dust (SPM) and its components
• hydrocarbons and chlorinated hydrocarbons
• low volatile organic materials (dioxins, furans, polychlorinated biphenyls).

A series of substances measured in this manner have been classified as carcinogens, such as cadmium compounds, PAHs or benzene. Monitoring them is therefore particularly important.

To provide an example of a comprehensive programme, table 7 summarizes the air quality monitoring that is systematically conducted in North Rhine-Westphalia, which with 18 million inhabitants is the most populous state in Germany.

Table 7. Systematic air quality monitoring in North-Rhine-Westphalia (Germany)

Back