Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community for help and clear up your study doubts
Discover the best universities in your country according to Docsity users
Free resources
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
UK Ordinance Survey. PM10. Fraction of airborne particulate that passes through a size selective inlet with a 50 % collection efficiency at ...
Typology: Lecture notes
1 / 89
This page cannot be seen from the preview
Don't miss anything!
May 2007
David Green, Timothy Baker and Gary Fuller
Environmental Research Group
King’s College London
Title King’s College London Volatile Correction Model for PM 10
Customer Department for Environment, Food and Rural Affairs (Defra), Scottish Executive, Welsh Assembly Government and the DoE in Northern Ireland
Customer Ref
File Reference ERG\Airquali\FDMS\Equivalence\Report\ KCL Volatile Correction Model for PM 10
Report Number KCLERG\MT\FDMS\EQ
Environmental Research Group King's College London 4th Floor Franklin-Wilkins Building 150 Stamford St London SE1 9NN Tel 020 7848 4044 Fax 020 7848 4045
Name Signature Date
Principal Author David Green^ May 2007
Reviewed by Gary Fuller May 2007
Approved by Gary Fuller May 2007
This report details the derivation, testing and application of a model to correct measurements of PM 10 by the Tapered Element Oscillating Microbalance (TEOM) such that they can be used to measure PM 10 for assessment against the EU Limit Value in the UK.
The First Daughter Directive (1999/30/EC) set Limit Values for PM 10 and also stipulated that PM 10 should be measured gravimetrically as laid out in EN12341 (CEN, 1998). There is however a conflict between the requirement to measure PM 10 gravimetrically^ and^ the requirement for rapid public reporting, and many member states, including the UK, rely on non- gravimetric techniques to measure PM 10. In the UK the majority of PM 10 measurements are made using the TEOM. In 2006, the UK PM 10 Equivalence Programme showed that the TEOM did not meet the equivalence criteria. The implied need to upgrade or replace TEOMs with an equivalent automated measurement technique has significant cost implications for DEFRA, the Devolved Administrations and for local authorities.
The KCL Volatile Correction Model (VCM) was based on analysis of daily mean measurements of PM 10 by Filter Dynamics Measurement System (FDMS) and TEOM at sites in the UK Equivalence Programme and at sites in the London Air Quality Network (LAQN). The model used the FDMS purge measurement (a measure of the volatile component of PM 10 , which is expressed as a negative concentration) to correct for differences in the sensitivity to volatile PM 10 between the TEOM and the EU reference method. The model equation for the correction of TEOM PM 10 measurements is:
Reference Equivalent PM 10 = TEOM – 1.87 FDMS purge
The FDMS purge concentration may be measured at a remote site, allowing the possibility of using a single FDMS instrument to correct the PM 10 measurements made by several TEOMs in a defined geographical area.
Three data analysis ‘experiments’ were undertaken to test the equivalence of the model to the EU reference method and to determine the maximum distance over which a FDMS purge measurement could correct measurements from TEOM instruments. The model was assessed against the criteria for the EC Working Group’s Guidance for the Demonstration of Equivalence of Ambient Air Monitoring Methods: the ‘Guidance’. Overall 772 equivalence tests were undertaken. The model passed the Guidance equivalence criteria at the sites used in the UK Equivalence Programme and can therefore be considered an equivalent method.
Further, the model passed the equivalence criteria using remote FDMS purge measurements over a maximum distance of approximately 200 km (22 out of 23 tests at less than 200 km passed the equivalence criteria, the single failure was marginal). This proves that the model is a viable tool for correcting measurements from TEOM instruments on the national and local government networks using FDMS purge measurements from a more limited network of sites.
The FDMS and the model were also tested against measurements at AURN sites. The AURN sites were outside the UK Equivalence Programme and employed slightly different measurement methodologies, although they did conform to the relevant CEN standard for PM 10 measurement. In most circumstances, the FDMS failed the equivalence criteria when tested against the AURN measurements. The model was derived from FDMS measurements and therefore unsurprisingly it also failed in many of these circumstances.
It was concluded that the failure of the FDMS and the model to pass the equivalence criteria at these sites was primarily due to the differences in measurement methodology at the AURN sites from that employed in the UK Equivalence Programme. For this reason, it is recommended that the measurement methodology at the AURN sites is bought into line with that used during the UK Equivalence Programme as far as is practically possible.
To enable the model to be applied to the TEOM measurements made routinely in the UK the configuration of the TEOMs should be changed to be the same as that used in the UK
ATP Atmospheric Temperature and Pressure.
DEFRA Department for Environment Food and Rural Affairs.
Emfab Teflon-coated glass fibre filters.
EN 12341 Standard for the reference measurement of PM 10 concentration
Equivalence Equivalent PM 10 measurement method to EN12341 according to the criteria in the Guidance.
EU European Union.
FDMS Filter Dynamics Measurement System.
FDMS purge
Mass concentration obtained from the FDMS at 30oC with sample having passed through dryer and 4oC chilled filter. Confusingly, termed FDMS Reference by the manufacturer.
FMDS Base Mass concentration obtained from the FDMS at 30oC with sample having passed through a dryer.
Guidance, the
EC (2005). Demonstration of Equivalence of Ambient Air Monitoring Methods, EC Working group on Guidance for the Demonstration of Equivalence.
KCL King's College London.
Klienfiltergerrat PNS-X8 a European Reference sampler for the measurement of PM 10.
NPL National Physical Laboratory.
nbs Number of between sampler datapairs
nc-s Number of candidate against reference datapairs
nEC Number of Daily LV exceedences for the candidate method
nES Number of Daily LV exceedences for the reference method
MS SQL Microsoft Structured Query Language.
OS UK Ordinance Survey.
Fraction of airborne particulate that passes through a size selective inlet with a 50 % collection efficiency at an aerodynamic diameter of 10 μm.
Partisol 2025 A sampler for the measurement of PM 10. Harrison 2006 found this sampler to be equivalent to the EU reference method.
TEOM Tapered Element Oscillating Microbalance.
UBS Between sampler uncertainty.
Upurge Between FDMS purge uncertainty.
WCM Relative expanded uncertainty at the limit value.
A) Derivation of the KCL Volatile Correction Model.
B) The model was then tested in three experiments:
Experiment 1) The model was tested using collocated FDMS and TEOM measurements and compared to the reference method measurements from the UK Equivalence Programme. In this experiment sufficient measurements were available to subject the results to the full equivalence tests. The objective of the experiment was to prove the equivalence of the model excluding the regional aspects.
Experiment 2) The TEOM measurements made in the UK Equivalence Programme were ‘corrected’ using FDMS measurements from distant sites and compared to the reference method measurements. Again sufficient measurements were available to subject the results to the full equivalence tests. The experiment had two objectives; to prove the equivalence of the model including the regional aspects and to begin to determine the model’s spatial applicability.
Experiment 3) The TEOM measurements at other UK sites were ‘corrected’ using FDMS measurements from distant sites and compared to the Partisol measurements. Results were judged in terms of the equivalence criteria. This experiment aimed to further determine the model’s spatial applicability.
C) The results from B) were used to determine the required spatial distribution of FDMS instruments to enable the model to be applied to all AURN TEOMs.
This section details the measurement methods used, the measurement programmes that supplied data, model derivation, statistical comparisons used, the coding of the model and the design of the experiments.
2.1 Measurement Methods
Four methods were used for measurement of the mass concentration of PM 10 or chemical speciation; these methods are described in sections 2.1.1^ to^ 2.1.4. Given the purpose of the study, the TEOM and FDMS methods are described in greater detail.
2.1.1 Gravimetric Measurement of PM 10
The gravimetric method for the measurement of PM 10 mass concentration forms the basis of the European and US gravimetric reference methods when used in association with defined operating parameters governing the choice of sampler, filter and method of laboratory (EPA, 1997; CEN, 1998; CEN, 2003); these are summarised in Table 1. The gravimetric samplers collected particulate matter onto a pre-weighed filter. The filter was then re-weighed under standardised conditions to determine the mass of particulate collected on the filter. Using measurements of sample volume, a mass concentration of particulate matter in the air was calculated.
EU PM 10 EU PM2.5 US EPA PM 10 US EPA PM2. Sampling Period 24 h 24 h (±1 h) 24 h 24 h Filter Media Quartz Fibre (^) Fibre, PTFE, EmfabQuartz Fibre, Glass Not Specified PTFE Filter Conditioning Temperature 20 ºC (±1 ºC)^ 20 ºC (±1 ºC)^ 15-30 ºC (±3 ºC)^ 15-30 ºC (±3 ºC) Filter Conditioning Humidity 50% RH (±3% RH)^ 50% RH (±5% RH)^ 20-45% RH (±5% RH)^
20-45% RH (±5% RH) Minimum Filter Equilibration Time 48 h^ 48 h^ 24 h^ 24 h Reporting Conditions 0 ºC, 101.3 kPa Ambient Ambient Ambient
Table 1: Specifications for the EU and US EPA Reference Methods
Two gravimetric sampling methods were used to measure PM 10 in this study, the Leckel Klienfiltergerrat PNS-X8 (KFG) and the Partisol 2025. The KFG is a European reference sampler for the measurement of PM 10. The Partisol 2025 was found to be equivalent to the reference sampler for the measurement of PM 10 in the UK using Teflon-coated glass fibre (Emfab) filters (Harrision 2006). The Partsiol 2025 was used as the reference sampler in the French Equivalence Programme, using quartz filters (Ampe et al ., 2005). Quartz filters are required for the measurement of PM 10 under the current legislation (CEN, 1998). However, Brown et al. (2006) found that the choice of filter media could impact on the repeatability of measurement due to the loss of filter integrity, the loss of sampled material during storage and transport and the degree to which of temperature and humidity during conditioning affect the mass measurement. It was expected at the onset of the UK Equivalence Trials, that Emfab would be included in the revised PM 10 standard as they are included in EN14907. Consequently, the Equivalence Programme used Emfab filters, while the AURN used (and continues to use) quartz fibre filters, in accordance with EN12341. Stricter protocols for storage, transport, conditioning and weighing were also used in the Equivalence Programme as proposed in Brown et al. (2006) and in EN14907. These included chilled storage and transport, additional conditioning time pre and post exposure, tighter temperature and relative humidity controls and the reweighing of unloaded and loaded filters to ensure repeatability (leading to discarding of filters).
2.1.2 Tapered Element Oscillating Microbalance (TEOM)
The TEOM is a real time particulate mass monitor, its mass measurement method relies on a microbalance, which consists of a hollow glass tapered tube, clamped at one end and free to oscillate at the other; an exchangeable filter is placed on the free end. The frequency of
accounted for the vaporisation from the filter measured during the purge cycle. A similar relationship has been found at Marylebone Road in London (Green, 2004) and is shown in Figure 1.
Figure 1: Time series of R&P 8400N daily mean ammonium nitrate measurements and daily mean FDMS purge measurements made at Marylebone Road during 2004
When examined at an hourly time resolution the relationship between the two metrics becomes increasingly clear. Figure 2 shows the hourly FDMS purge and ammonium nitrate measurements at Marylebone Road during April 2004. This month was chosen because the relatively low temperatures and prevalence of long range transport of secondary particulate matter during the spring leads to some of the highest particulate ammonium nitrate concentrations found in London each year. The FDMS purge and ammonium nitrate measurements shown in Figure 2 correlate fairly well (r^2 = 0.50). However a closer examination reveals a small lag between the peak concentration of ammonium nitrate and the maximum FDMS purge measurement. An example is shown on the inset graph in Figure 2, which highlights the hourly measurements from Marylebone Road between 16th^ April and 18th^ April (the highest concentrations during that month). The dotted line shows the time of the maximum ammonium nitrate concentration; 23 μg m-3^ at 6:00 am on 17th^ April. The maximum FDMS purge concentration was measured one hour later. This was consistent with previous studies, which showed that the lag time between the ammonium nitrate concentration measurement and vaporisation was between 40 minutes and 100 minutes (Hering et al. , 2004). Indeed, if the time of the FDMS purge measurements during April is lagged by two hours, the r^2 increases to 0.62.
Figure 2: Time series of R&P 8400N hourly mean ammonium nitrate measurements and hourly mean FDMS purge measurements made at Marylebone Road during April 2004. The inset graph highlights the peak concentrations (shaded) measured on 17th^ April.
Orthogonal regression analysis of the more limited collocated ammonium nitrate and FDMS purge measurements at Harwell and Belfast are shown in Figure 3 alongside those from Marylebone Road. Strong correlations were found, however, slopes varied between 0. (Harwell) to 1.99 (Belfast), the Marylebone Road slope was 1.08. All the intercepts were close to zero. The relationship between the FDMS purge measurement and ammonium nitrate concentration therefore appears to vary at different locations where it is measured in the UK. However, as discussed a near 1:1 relationship has been demonstrated in London and at several sites in the USA and the reason for this differing relationship at other sites in the UK clearly requires further investigation.
Figure 3: Orthogonal regression analysis between the NH 4 NO 3 concentration in PM2.5 and the FDMS purge measurements at Marylebone Road between 17th^ February 2004 and 31st^ December 2005.
Figure 5: A schematic of the FDMS system. Base cycle (top) and purge cycle (bottom) configurations are shown separately.
2.1.4 Nitrate measurement
Nitrate measurements were made using the Rupprecht & Patashnick (R&P) Nitrate Monitor 8400N. The R&P 8400N is a near real-time particulate nitrate monitor, measuring the inorganic nitrate composition of PM2.5. It consists of two instruments: a C3 pulse generator and a NOX pulse analyser. The instrument was operated at default settings, although the cycle time was set to 15 minutes so that it could be directly compared to the TEOM instruments (before May 2005 at Harwell and Belfast this was 10 minutes). Every 15 (or 10) minutes, collected particles were flash-volatilised in a nitrogen atmosphere by resistive heating of the NiChrome strip. The NOX pulse analyser measured the resulting pulse of NOX.
2.2 Measurement Programmes
Measurements were obtained from four measurement programmes.
2.2.1 UK Equivalence Programme
The Defra funded UK Equivalence Programme (Harrison 2006) was a bespoke measurement programme designed to test the equivalence of seven candidate instruments to the EU reference methods for the measurement of PM 10 and PM2.5 concentration. The programme was managed by Bureau Veritas and included the operation of instruments at four locations in the UK; Teddington (suburban London), Bristol, Birmingham and East Kilbride. Measurements from the four locations were divided into separate summer and winter deployments to provide eight field campaigns from late 2004 to early 2006.
The candidate instruments included paired TEOM and FDMS for the measurement of PM 10 , as well as other beta attenuation and sampler instruments not used in this study. The reference instrument for PM 10 measurement was the KFG. Filters were changed daily at 10h, 11h or 15h GMT.
Gravimetric measurements were made using Teflon-bonded glass fibre (Emfab) (Pall Corp., NY, USA; Type: EMFAB TX40HI20-WW; Part No.: 7221). As discussed in section 2.1.1, Emfab filters were chosen after extensive investigation of the properties of various filter media by the National Physical Laboratory (NPL)(Brown et al. , 2006).
NPL were responsible for the provision of quality assured gravimetric PM measurements and AEA Energy and Environment undertook audits of each reference and candidate instrument. Prior to the full equivalence trial FDMS instruments were also briefly deployed at Belfast (urban centre) and Harwell (rural).
All measurements from the UK Equivalence Programme have been made available online and can were obtained from:
www.airquality.co.uk/archive/reports/cat05/0607131442_UK_Equivalence_Trials_Data.xls
and entered into the KCL air quality database.
2.2.2 The London Air Quality Network
The London Air Quality Network (LAQN) was formed in 1993 and comprises of over 100 local authority funded monitoring sites in London and the Home Counties. The network is managed by KCL.
During 2003 KCL instigated a FDMS monitoring programme with the following objectives:
each midnight and filter weighing was carried out by Bureau Veritas according to the conditions in EN12341(CEN, 1998). Sample flow rates were subject to UKAS accredited audits and measurements were ratified by AEA Energy and Environment. Measurements were obtained from www.airquality.co.uk. TEOM measurements were processed to create 24-hour mean concentrations commensurate with gravimetric filter change times. A 90% hourly data capture was required to create a valid 24-hour mean concentration. Temperature and pressure measurements for the conversion of reported TEOM measurements to ambient temperature and pressure were obtained from the Belfast City Council for the Belfast site and Hertfordshire and Bedfordshire Air Pollution Monitoring network site at Luton, Bedfordshire, UK (OS 506500 222700, Lat: 51.892687N Long: 0.453770W, Datum: WGS84) for the other sites.
2.2.4 Other Defra Gravimetric Measurements
Other gravimetric measurements were also undertaken using the Partisol 2025 onto quartz filters. Filter handling and weighing was identical to that described in section 2.2.3. Sample flow rates were subject to UKAS accredited audits by AEA Energy and Environment and measurements were ratified by Bureau Veritas. Measurements were kindly supplied by David Harrison.
2.2.5 Defra Airborne Particle Concentration and Numbers Network
The Airborne Particle Concentration and Numbers Network measures concentrations of particle nitrate, sulphate, chloride, carbon and particle numbers at eight urban and one rural site in the UK. Measurements of nitrate in PM2.5 were made at three sites as listed in Table 2. From May 2005, operational management of these instruments was undertaken by KCL with audit and the measurements ratification undertaken by NPL. Prior to May 2005 site management and ratification of the R&P 8400N at the Marylebone Road site was undertaken by KCL and Bureau Veritas undertook these responsibilities for the remaining instruments and sites.
2.2.6 Drax Power Measurements
Two FDMS instruments were operated by Drax Power Ltd around their power station in Yorkshire (Anwyl, 2005). The instruments were subject to 3 monthly flow calibration. Measurements were collected and fault checked by Phil Anwyl of AQMS Ltd. Measurements were placed in the KCL database and subject to screening ratification checks.
Figure 7: FDMS and Partisol sampling locations used in this study in UK, further detail of the London sites is shown in Figure 6.
