You are here: Publications > Air quality >

Ambient Air Quality Monitoring and Data Management Workshop report - monitoring ambient air quality

16 March 2001

The main points raised during the morning presentations on ambient air quality monitoring are detailed below. For more information, refer also to the individual presentations.

Welcome and introduction
Deciding on an air quality monitoring programme
Setting up a monitoring site
Which methods to use
Standard and non-standard methods and issues not covered by the Guide
Industry Perspective on the Good Practice Guide
Identification of Ambient Air Quality Monitoring Issues - Panel and Group Discussion

Welcome and introduction

Caroline Austwick, MfE.

The Good Practice Guide (GPG) aims to provide assistance to enable interested parties to achieve high quality, cost effective ambient air monitoring using nationwide consistent methods (when appropriate).

Deciding on an air quality monitoring programme

Paul Sheldon, Nelson City Council.

Ambient air quality monitoring is an iterative process that may develop in parallel with the regional plan but under the umbrella of a long-term goal.
Getting out and about and using local knowledge is the one of the key elements to a successful monitoring programme.
Differentiate between national SOE monitoring and local effects. One monitoring programme may not meet both needs.

Setting up a monitoring site

Vera Hally, Air and Environment Science Ltd.

Gave practical advice on where and how to set up a monitoring site.
It is vital that processes and actions are identified, documented, and labelled. Put this record somewhere where people can find it.
Don't forget OSH requirements especially when handling compressed and/or explosive gases.

Which methods to use

Emily Wilton, Environment Canterbury.

Presentation focused on PM₁₀ sampling. Which black box do you use?
Don't unquestioningly believe the sales pitch.
A lot of money does not necessarily buy freedom from problems or avoid the need to understand what is going on inside the black box.
Look carefully at what you want to measure. Understand how you are going to measure it.
Hi-volume verses TEOM PM₁₀ monitoring can shows a difference of about 20 %.
There are significant issues to be aware of if employing optical sampling methods.

Standard and non-standard methods and issues not covered by the Guide

Gavin Fisher, NIWA.

Should calibration methods be prescriptive?
How do you deal with outliers?
New standards are currently being written in Australia.

Industry Perspective on the Good Practice Guide

Paul Barrett, Ravensdown Co-operative Ltd.

Applying the GPG to situations increases its value. Praised practicality of Guide.
Suggested that some additional information on cost implications of the different monitoring options would be useful.
Raised question of what contribution industry should be required to make to ambient air monitoring programmes.
Would like to see criteria that provide guidance on what contaminants to monitor.
Should information from industrial monitoring programmes go onto the Air Quality database? Strong positive response to this question.
Mentioned usefulness of control sites.
There are different requirements of impact (from an industry site) and ambient monitoring.

Identification of Ambient Air Quality Monitoring Issues - Panel and Group Discussion

Chaired by Gary Bedford, Taranaki Regional Council

Key issues identified by Gary and discussed by panel and floor.

1) Which contaminants should be measured?

Establish if there is a (potential) problem. If there is a potential problem then it is essential to have a long-term strategy (5, 10 and 15 years) which gives guidance on what pollutants to measure and when.

To decide what will be monitored it is suggested that it is useful to consider:

(Revised) Ambient Air Quality (AAQ) Guidelines;
The site specific relevant environmental consequences;
Indicators or the effectiveness of policy (yet to be developed!);
Emission inventories;
Inference from co-contaminants e.g. PM₁₀ and SO₂ from coal combustion;
Results of screening air quality programmes using methods such as passive samplers;
Use local knowledge e.g. visible smoke can be an indicator of PM₁₀ and SO₂;
Compare your site with similar sites where some air quality data already exists;
Ask others with experience in the field.

Public expectation may be created by the GPG that all pollutants will be monitored.

How much reliance can be placed on current (revised) AAQ guidelines to identify the pollutants that require measurement? AAQ guidelines can be used as a political tool to gain funding for monitoring programmes.

Industrial perspective on which contaminants should be measured.

When a new plant is planned who has the responsibility of establishing the background levels? It was suggested that Regional Councils should provide regional data. Other members of the workshop pointed out that industry was responsible for establishing background levels, where no existing data is available, as this allowed the assessment of cumulative effects (an RMA requirement). Another point of view is that there is a statutory requirement under the RMA that industry is responsible for monitoring their impact. This would imply that industry is responsible for establishing baseline data. Some inconsistencies between different Regional Councils were noted on this matter.

Consent conditions can require that monitoring be carried out but these tend to require more stack testing than ambient levels. Different consent requirements are made from region to region. Is some sort of national consistency required? Some danger in this approach and an assessment on a case by case level is probably the most appropriate method.

In addition, what drives the decision whether monitoring is undertaken, fear or science? Public concern often a greater issue than a scientific effects based assessment. In this case complaints records are very useful tool. Ravensdown Fertilizer Hornby, Christchurch monitors for SO₂ to keep public happy.

2) Frequency of monitoring

Not all monitoring programmes need to be continuous or repetitive. Short-term monitoring programmes also have their place. Short-term monitoring programmes are especially useful for establishing how detailed long-term monitoring needs to be and for showing the public that something is being done.

It was suggested that a minimum of three months is a realistic timeline to attain quality data and get a good feel for the magnitude of events. However a three month period may only be appropriate when maximum emissions coincide with poor dispersion. Choose the timing of your monitoring campaign carefully. Otherwise to cover the complete range of meteorological conditions that occur a much longer period (of two years?) may be appropriate.

Councils suggested that having a mobile trailer (or moveable shed) set up to measure a variety of pollutants and that is available for councils to lease would be very helpful.

3) Representativeness of sampling.

Fine scale meteorology often determines what you will measure (e.g. exposed vs sheltered areas). Watch for setting monitoring sites up in pollution traps that may give higher ambient levels than are representative of the area under study. Whole monitoring programmes can be a waste of resources if the monitors are placed in the wrong location. Ensure that you ask all the right questions before establishing the locations at which pollutant levels are to be monitored.

Regional pollution hot spots can be identified by either mobile monitoring, modelling or a simple (and relatively cheap) screening programme. Passive sampling is good for long-term trends and for checking results of emission inventories but does not give any indication of fluctuations of short-term levels.

Once the data is gathered it must be analysed to decide if the results show worst case scenarios or typical exposure i.e. are the short term peaks caught by the monitoring or is the long term cumulative effect being observed?

4) The importance of meteorological information

Having meteorological data is essential. Meteorological data adds significant value to the monitoring data by allowing a more in depth analysis of monitoring data to occur thus helping to:

Establish the representativeness of monitoring data;
Explain why events occur;
Identify sources of pollutants;
Identify potential mitigation measures.

The importance of meteorological data needs to be strongly emphasised. May be wasting monitoring resources without collecting good meteorological data. To date the value of meteorological data has been underestimated.

Data from closest meteorological station can potentially be used. Much of this data is available on the NIWA Climate database (CLIDB) or can be obtained from other sources e.g. councils, Metservice and individual operators. The relevance of this data to your site will be often determined by fine scale meteorology. Fine scale meteorology will reflect local topography and in New Zealand this is often complex. An assessment must be made of the relevance of the closest meteorological data source to the area under investigation.

Models can be used to provide meteorological information. But models have large (and sometimes expensive) data requirements. Modelling can be just as expensive as using real meteorological monitoring gear. Some models uncertainty increases as wind speeds decrease. This is a problem if high pollution events occur during periods of low wind speeds. Modelled information is only useful if you understand the model's limitations. It was also noted that the modelling of events was a much easier thing to achieve than establishing long-term ambient levels via modelling.

It was suggested that meteorological data should be the primary information to be collected. There was some disagreement amongst the group on this matter. Some parties suggested it is much more useful to have both concentrations and met data.

Met data is especially important if you are concerned with short-term events rather than longer term averaging periods. This is because the relationship between short-term pollutant concentrations and meteorology is more complex.

It was suggested that at least two years of meteorological data was required from a site to account for extreme events, seasonal and annual variation. It was pointed out that collecting this amount of data would have large financial implications.

5) Cost and cost effectiveness of air quality monitoring

Something you must choose between is a lot of low cost, low quality monitoring sites (screening study) and a few sites providing high cost, high quality data (that may not give you a region wide perspective). Determine what quality of data is required before starting monitoring campaign so that you can get the best "bang for your buck".

Because of the high set up costs for each site, the potential flexibility gained by site to site movement of instrumentation is low. Some flexibility can be gained by using mobile trailers and/or relocatable sheds. It was noted that relocatable sheds are normally cheaper than trailers.

6) "Future Proofing" of GPG

What can we do to ensure that this guide can do all it can to provide guidance on "surprise" issues? e.g. Benzene following the banning of leaded petrol and the PM_2.5 issue. Some councils already have an interest in monitoring PM_2.5.

There is a move from assessing PM impacts by a mass concentration to a particle size distribution on a risk assessment basis. What implications would this have for the use of the GPG?

We need to be able to look outside the box and be able to anticipate problems. Emission inventories can provide useful information in this area. Must have an awareness of impending documents and guidelines. Networking with other agencies that undertake monitoring will be required. May be more difficult for smaller Councils.

7) Other air quality monitoring issues not addressed in the GPG

Remember time delay from planning monitoring programme and getting good quality data e.g. delivery of calibration gases.
Some contaminants need "specialist" monitoring e.g. dioxins, PAHs and odour.
Moving monitoring sites may mean some down-time in monitoring record. Ideal situation would be to have sites overlap in temporal coverage.
There are financial advantages to using a relocatable shed over a mobile trailer.
Power supply often a major issue when selecting monitoring sites. On site generation can be an option (e.g. solar panels, but not diesel generators).
Some analysers can take time to stabilise. It was suggested this could take up to three months.
Is there a need for wider ozone monitoring? Region by region decision.
Is more information on running and maintaining analysers needed?
Procedures for the use and maintenance of regulators, cylinders etc.
Should the limitations/constraints of monitoring methods be stressed more strongly GPG? e.g. Zero air can contain up to 0.5 ppm CO.
When can it be more appropriate to use source monitoring?

8) Relationship between science, policy and monitoring AAQ

Some generic questions on the use of science in policy making were raised.

Do you need to wait for the science before you bring in the policy?
How much scientific proof do you need to justify policy?

It was generally agreed that a well-balanced multi-pronged approach is most effective. This is common ground with the management of many resources eg air, soil and water.

Monitoring provides scientific evidence to justify methods/policies used. If science is discredited then the policy is likely to be ineffective.

Monitoring results can be used as a publicity and community relations tool to get support to action policy. But must avoid brain washing and propaganda! Do not underestimate the power and usefulness of the media. A good deal of caution must be used in this forum. Summary documents backed up by technical reports are a useful approach.

Where should the financial emphasis be put, Science or Policy? The floor suggested that a starting point is $10:$1 science:policy. Some planners may not agree!

VFECs programme has approximately $10M of data behind it. It now requires environmental monitoring to turn this into an effective planning tool.

There is an implicit requirement in the RMA that monitoring is needed to assess whether policy is effective.

There is a feedback loop between monitoring and policy which could be something like:

Science/Data Transfer => Knowledge => Policy => Assess Policy effectiveness => feedback to Science/data transfer.

Is this feedback loop being used effectively in New Zealand? AAQ monitoring plays an important part in this. Should we attempt to integrate all GPG and AAQ guidelines into this feedback loop?

9) Additional comments from participants

Vera Hally, Air and Environmental Science Ltd.:

"much of the confusion relating to negative values relates to understanding the limitations of the measurement equipment involved and understanding some fundamental principles of measurement".

"It is misleading, and potentially dangerous, to state that calibrations cannot be performed using permeation tubes, and only with compressed gases. Again, there are management protocols for both methods which determine how good that data/information is. Refer to AS 3580.2.1 (AS2522) the preparation of reference test atmospheres - Permeation tube method, and AS 3580.2.2 (AS2573) the preparation of reference test atmospheres - Compressed gas method. Both have the potential to go wrong if the critical managment principles are not recognised and adhered to. By stating that "permeation tubes are not OK, but compressed gas is", you minimise and trivialise the importance of understanding and managing these issues to the degree of absolutely ignoring them!"

Andrew Harper, NIWA

Had a few points, mainly relating from a meteorological perspective.

It is not clear what height should be used for temperature sensors and what type of solar radiation should be measured. Is that global solar radiation or net radiation? These are two quite different things.
There is no mention of calibration and servicing of meteorological sensors (should be annual or bi-annual depending on sensors).
The document mentions a 1:10 height ratio for siting. This is very difficult to achieve. The WMO uses 1:2 for most parameters.
The document mentions rainfall to a resolution of 1mm. Standard raingauge tips are typically 0.2mm or 0.5mm.
If the meteorological data is to be comparable with other meteorological monitoring stations, then it is important that the above factors are taken into account to ensure consistency.