Monitoring of Ambient Air Particulate Arsenic ...

Monitoring of Ambient Air Particulate Arsenic Concentrations at Nine Sites in Cornwall

Final Report December 2006 Jo Barnes, Alex Ledbrooke, Barbara Parsons and Leo Salter Air Quality Unit Cornwall College [email protected]

Monitoring of ambient air particulate arsenic concentrations at nine sites in Cornwall Final report 2006 1.0 Introduction Although arsenic is a human carcinogen (DEFRA, 2002), there are currently no UK air quality standards for particulate arsenic in air, even though the World Health Organisation (WHO) (1987) acknowledges that there is no safe limit to be set. The health risk assessments which have been made report a unit risk value of 1.5 x 10-3 for each µg m-3 of arsenic to which an individual is exposed (WHO 1987) and WHO (2000) also states that “When assuming a linear dose-response relationship, a safe level for inhalation exposure cannot be recommended. At an air concentration of 1 µg m-3, an estimate of lifetime risk is 1.5 × 10–3. This means that the excess lifetime risk level is 1:100,000 at an air concentration of about 6.6 ng m-3”. In the UK airborne arsenic can, on occasion, be problematic. For instance, in the DEFRA report “A Review of Arsenic in Ambient Air in the UK” (Maggs 2000) annual mean concentrations of airborne arsenic in the UK rural environment for 2000 were found to be typically in the range of 1 – 4 ng m-3, whilst annual mean urban concentrations were higher, in the range of 5 – 7 ng m-3. The report states that the highest UK concentrations of airborne arsenic are typically found near industrial sites and quotes the highest annual mean concentration of arsenic (found in the vicinity of a smelter in Walsall) for 1975 – 1990 as 223 ng m-3. Concentrations of arsenic at other urban sites, during the same period had annual mean values which varied from 1 – 18 ng m-3. Some of these values exceed those proposed by the EC Directive 2004/107/EC (EC 2004) relating to arsenic in ambient air, where their recommendations are that “. . . mandatory monitoring should be introduced where concentrations in ambient air exceed thresholds of 6 ng m-3 for arsenic.” At concentrations below those levels, where the Commission says that the harmful effects on human health would be minimal, only indicative monitoring would be required with a limited number of sites targeted (CEC 2003). The Directive also comments that “. . . the oral uptake of arsenic is the most important route of exposure, however, regarding its carcinogenic effect, inhalation is of major importance.” Maggs (2000) also states that with regard to urban concentrations of airborne arsenic, that domestic coal burning has historically contributed to arsenic-in-air concentrations but that this source is declining due to recent shifts in the number of households in the UK burning coal. 1.1 Arsenic in Cornwall In the UK, Cornwall’s history of mining and smelting has left a legacy of contamination and pollution. Between 1860 and 1900, Cornwall was the world’s major producer of arsenic and at the height of the industry also imported arsenical ores for smelting (Kavanagh et al. 1998). As a result of this activity soil in some parts of Cornwall has the world's highest concentration of arsenic (up to 2500 ppm) (Xu and Thornton 1985). Elsewhere in the UK and the rest of the world soil arsenic concentrations are usually below 40 ppm (WHO 2003). Previous studies looking at residents of Cornwall living in former mining areas found concentrations of arsenic in their urine were greater than that of those in the control group, demonstrating that aspects of arsenic effects on the Cornish population are of concern (Kavanagh et al. 1998).

AQU: Arsenic in ambient air Final report December 2006

1

This programme focuses on concerns in Cornwall about human exposure through ingestion or inhalation of arsenic-containing airborne dust and soil particulates. Some data in Cornwall already exists and is described below. Environment Agency (EA) data for arsenic concentrations in soil and dust samples collected from the United Downs site (Figure 1) and from roadside sites in 2002, found elevated values of arsenic consistent with those previously reported for mineralised areas in Cornwall (Culbard and Johnson 1984) and for arsenic concentrations in stream sediments in Devon and Cornwall (EA in preparation; Robins pers. comm.).

Figure 1: United Downs In 2003 the Air Quality Unit (AQU) undertook a dust (PM10) monitoring programme on behalf of Kerrier DC and Carrick DC. This work stemmed from public complaints received by the Local Authorities from residents living in the vicinity of United Downs landfill site. As an aside to the original objective of the monitoring programme (recording PM10 concentrations) the AQU had filters analysed to determine the arsenic concentrations present in the PM10 fraction of airborne matter. This preliminary work revealed arsenic concentrations (in the PM10 fraction) between 2 and 13 ng m-3 (CAQF 2003). Concentrations recorded at United Downs were elevated compared to those found in other UK rural areas (0.8 – 4.5 ng m-3) but within the range of concentrations recorded at urban sites (1 – 18 ng m-3) (Maggs 2000). Due to the lack of historical monitoring data regarding concentrations of airborne particulate arsenic in Cornwall there is no way of assessing whether the data gathered in 2003 shows an increase or decrease in airborne particulate arsenic concentrations. Although data related to arsenic in ambient airborne particulate matter in the UK is available from NETCEN (http://www.airquality.co.uk/archive/) there are no monitored sites in Cornwall. Without monitored data relating to background concentrations of ambient particulate arsenic, it is not clear whether the concentrations monitored at United Downs were temporarily elevated or the local norm. The aim of this monitoring programme is to provide more robust data concerning the baseline levels of particulate arsenic in ambient air in one of the most heavily arsenic contaminated areas in the UK. The data will provide a benchmark against which future arsenic air quality monitoring in Cornwall and other areas of the UK can be assessed.


2

1.2 Particulate matter (PM10) Airborne particulate matter is classified by size and the size categories are directly linked to the distance particles can penetrate into the lung; the smaller the particle the further it will penetrate. PM10 is particulate matter below 10 micron aerodynamic diameter and is designated as the upper limit of the respirable fraction, that is, the mass fraction of inhaled particles which penetrate to the unciliated airways of the lung and as such, are thought to be a major health hazard. With regard to monitoring equipment used in this programme, “PM10” means particulate matter, which passes through a size-selective inlet as defined in ISO 7708:1995 with a 50 % efficiency cut-off at 10 µm aerodynamic diameter. The European Union (EU) Position Paper “Ambient air pollution by As, Cd and Ni compounds” makes the point that because arsenic compounds are enriched in the fine particle mode (≤ 1µm) they will penetrate deeply into the respiratory system (EU 2001). Under normal conditions the oral uptake of arsenic is generally the most important route of exposure, whereas inhalation normally contributes less than 1 % to the total dose. Noncancer effects observed after inhalation of air with high arsenic levels at workplaces are increased mortality from cardiovascular diseases, neuropathy and gangrene of the extremities. There is also sufficient evidence to demonstrate that inorganic arsenic compounds are skin and lung carcinogens in humans. At present the possibility cannot be ruled out that any form of inorganic arsenic may be carcinogenic. Consequently, lung cancer can be considered to be the critical effect following inhalation exposure (EU 2001). The sampling methods utilised in this monitoring programme collected PM10 samples using the Partisol and Osiris monitor and Total Suspended Particles (TSP) samples using the DustScan monitor.

2.0 AQU 2005 – 2006 airborne particulate arsenic monitoring 2.1 Sites To provide an overview of ambient airborne arsenic concentrations in Cornwall and focusing particularly on the contribution that areas of traditional mining processes may make to those concentrations, nine sampling sites were chosen across a 250 km2 area, covering a range of past mining activities (Figure 1). The sites were chosen with specific reference to four criteria: • • • •

Relevance to residential properties The proximity to historic mining sites Open aspect with exposures to all wind directions The provision of a power supply and security

Four monitoring sites were located within 500 m of historic mining sites; these were Porthtowan, Bissoe, Camborne and Carharrack. The site at Crowan was within 1 km of mining sites, Falmouth and Constantine within 2 km and Trelissick was chosen as a background site as it had no mining sites within a 4 km radius (Figure 2). Stithians was designated as a control site and samples were taken at this site throughout the whole of the monitoring period. Stithians was chosen as the control site mainly because it was central to the sampling area, but also because it was possible to locate monitoring equipment at the site for an extended period. It was anticipated that samples taken at Stithians for each monitoring period would provide a continuous record of ambient airborne arsenic concentrations which would enable comparisons to be made with concentrations at other sites. Due to equipment restrictions it AQU: Arsenic in ambient air Final report December 2006

3

was not possible to monitor at all sites simultaneously. Three sites, including the control site, were therefore monitored sequentially over four programmes. The monitoring programme is detailed in Tables 1 and 2 (Appendix 1).

© Crown copyright. All rights reserved. 100019590. 2006

Figure 2: Nine monitoring sites (labelled) and sites of historic mining (small dots). Buffers show the distance of monitoring sites to the nearest mine site. Mine site data supplied by the EA.


4

2.1.1

Plate 1: Bissoe

The site at Bissoe (SW775412) is located in the reed bed enclosure in the Bissoe Valley adjacent to a cycle path. The site is the property of Wheal Jane Ltd. The Point Mills arsenic refinery was located in the Bissoe valley and ceased operating in 1939 (World Heritage Cornwall, 2006). There are also a number of historic mining sites and sites of exposed mining spoil within 500 m of the monitoring site. The area is a popular tourist leisure area which is well frequented by walkers and cyclists.

2.1.2

Plate 2: Carharrack (Samsfield)

The Carharrack site (SW738413) is a rural site located approximately 0.5 km south west of the CES landfill site at St Day. The site is also close to the proposed landfill site extension at Arsenic Woods.

2.1.3

Plate 3: Constantine

The Constantine site (SW742310) is a rural site located in the grounds of a garden centre. The site is within 2 km of a mining site.


5

2.1.4

Plate 4: Crowan

The Crowan site (SW646345) is located in the garden of a residential property in a rural village within 1 km of mining sites.

2.1.5

Plate 5: Falmouth

The Falmouth site (SW805326) is an urban site which is within 2 km of mining sites. The monitoring equipment was situated on the roof of Falmouth Marine School.

2.1.6

Plate 6: Porthtowan

The Porthtowan site (SW692480) is located in the garden of a residential property adjacent to the beach. The monitoring site is within 500 m of mining sites and exposed mining spoil.


6

2.1.7

Plate 7: Stithians

The control site at Stithians (SW721366) is located centrally in relation to the other sites in order to minimise the effects of other variables, e.g. sea salt. The site was located in the garden of a residential property in a rural area approximately 3 km from mine sites.

2.1.8

Plate 8: Trelissick

The Trelissick site (SW837396) is located within a walled garden in the grounds of Trelissick House. This site was considered to be a background site in terms of arsenic exposure, being situated 4 km from a mining site.


7

2.1.9

Plate 9: Camborne (Tuckingmill)

The Tuckingmill site (SW660396) is located in the garden of a residential property in an urban area of Camborne. There are mine sites situated within 50 m of the monitoring station and exposed mine workings, e.g. South Crofty in the immediate vicinity.

Plate 10: South Crofty At the Tuckingmill site the South Crofty mine workings can be seen behind the houses opposite.


8

3.0 Equipment A range of airborne particulate monitors were used in this study to sample the ambient air for PM10, i.e. • • •

Partisol Model 2000 and inlet head DustScan Osiris particulate monitor (control site only)

Wind speed and direction data was obtained from a Solid State Wind Sensor MMW 005 sensor. 3.1 Partisol The Partisol Model 2000 Air Sampler (Rupprecht & Patashnick (R&P) Co. Inc.) (US EPA (United States Environmental Protection Act) equivalent method) is a microprocessorcontrolled manual sampler equipped with a PM10 inlet head (Plate 11). The unit contains the sample inlet (PM10) a 47 mm filter exchange unit, a microprocessor with internal data storage, an active flow control system and a pump. The Partisol draws 16.7 l min-1 (1 m3 h-1) of flow through the sampling head. The total flow then passes through the 47 mm diameter filter. This flow is equivalent to the breathing rate of a human adult. The Partisol has a SA246b PM10 inlet (Graseby Anderson) consisting of a flanged cylindrical pipe with a disc rain cap held 10 cm above. This forms the omnidirectional narrow slot entry through which particles enter, followed by a single stage multi-orifice impactor, which allows the PM10 fraction to penetrate to a filter. The sampling efficiency of the entry is in full agreement with the US EPA PM10 convention. Plate 11: A Partisol monitor with PM10 inlet head

The Partisol stores data relevant to each 24-hour period on an internal data logger. These data comprise the total flow volume (in terms of standard temperature and pressure), the total collection time and the average temperature during the collection period. The Partisol is serviced on a 6-monthly basis by Air Monitors Ltd. Three Partisol monitors were deployed in each programme; one at the control site (Stithians) and one at each of two other sites).


9

3.2 DustScan The DustScan consists of an adhesive dust collection sheet mounted on a cylinder on a 2 m high stand (Plate 12). The monitoring head collects fugitive Total Suspended Particulate (TSP) from 360° around the installation. The sheet is then retrieved for analysis to identify the direction from which deposition occurred. DustScan samples can be further analysed by QEMSCAN™ or Scanning Electron Microscopy (SEM). One DustScan monitor was co-deployed with each Partisol at each site with dust collection sheets exposed for the whole programme.

Plate 12: DustScan gauge in position (left) and Partisol (right) (Porthtowan)

3.3 Osiris Particulate Monitor

The Osiris particulate monitor (Turnkey Instruments) employs a light scattering technique which gives a continuous and simultaneous indication of the TSP, PM10, PM2.5, and PM1.0 mass fractions. The Osiris monitor is approximately 90 cm x 25 cm x 25 cm and consequently is easy to deploy. The Osiris was deployed at the control site (Stithians) only and monitored for the whole of each of the monitoring periods in 2005 and 2006.

Plate 13: The Osiris particulate monitor


10

3.4 Solid State Wind Sensor MMW 005

The solid state wind sensor uses a temperature sensitive element to measure wind direction and wind speed (Plate 14). The wind sensor was deployed at the control site (Stithians) only. Unfortunately, due to data logging failure, no wind data is available for the 2006 monitoring period.

Plate 14: Wind sensor 3.5 Microbalance Filters were massed before and after exposure using a Mettler Toledo: Mettler MT5 Electronic Digital Top Pan Microbalance weighing to 1 x 10-6 g (1 µg). The Mettler Toledo Microbalance is serviced yearly by Mettler Toledo and calibration checks are performed.

3.6 Temperature and humidity chamber

To condition the filters to the temperature and humidity specified by CEN (European Committee for Standardisation), before and after exposure, a temperature and humidity chamber was built. The chamber consists of an airtight clear PVC tank with access via a door and two glove ports. A CAL 3200 DIN process controller and a Rototronic Hygromec FT series humidity and temperature transmitter control the temperature and humidity (Plate 15 (with microbalance)).

Plate 15: The humidity chamber and micro balance


11

3.7 Filter type A number of filter media were tested to determine their suitability for airborne particulate arsenic monitoring (Table 1). Table 1: Filter media tested Pallflex Teflon-coated glass fibre Pallflex Tissuequartz Millipore PVC Whatman Nylon 0.45µm Whatman Nylon 0.80µm Whatman Cellulose Nitrate 0.45µm Whatman Cellulose Nitrate 0.80µm The type of filter was dictated by the arsenic analysis technique specifications. Whatman cellulose nitrate 0.8 µm filters were chosen as they have zero blank arsenic levels (for ICP-MS digestion) and do not produce volatiles under electron microscopy that might damage the QEMSCAN™. 3.7.1

Protocols for filter conditioning and weighing

In accordance with the protocols specified by CEN standard document prEN 12341, filters are conditioned at 50% ±3% humidity and 20° C ±1° C for a 48-hour period. This is an EC standard, as opposed to US EPA or UK DoE and in order to ensure full compatibility with other European member states the EC standard was strictly adhered to. When dealing with very small changes in mass the effect of temperature and humidity on a filter is significant and a relatively small change due to water vapour uptake can have a major effect on calculated concentrations. Careful regulation of the conditioning process to within the limits specified above is therefore essential. After conditioning, filters were loaded into the filter holders (Plate 16) and stored in clean, labelled Petri dish for transportation to and from the site. After the filter had been exposed for the allotted time it was removed from the Partisol filter exchange unit and placed back into its labelled Petri dish for transport to the laboratory – care was taken not to invert the filter. The Partisol’s internal data storage was then accessed and the average temperature, atmospheric pressure and standard flow volume recorded. The mass of particulate matter sampled was determined by weighing the filters used prior to and after the test period using the Mettler Toledo: Mettler Microbalance and deducting the post-exposure weight from the pre-exposure weight. Each filter was weighed three times and the results averaged to determine the mass of particulate matter sampled (ensuring a zero balance between weighing).

Plate 16: A clean filter loaded in a filter dish AQU: Arsenic in ambient air Final report December 2006

12

3.8 Filter analysis techniques 3.8.1

ICP-MS

Inductively-coupled plasma mass-spectrometry (ICP-MS) was used to identify metallic and metalloid elements on the filters. In ICP-MS, a plasma or gas consisting of ions, electrons and neutral particles is formed from argon gas. The plasma is used to atomize and ionize the elements in a sample. The resulting ions are then passed through a series of apertures (cones) into the high vacuum mass analyzer. The isotopes of the elements are identified by their massto-charge ratio (m/e) and the intensity of a specific peak in the mass spectrum is proportional to the amount of that isotope (element) in the original sample. 3.8.2

QEMSCAN™

QEMSCAN™ system is 1,000 to 10,000 times faster than traditional methods of quantitative analysis and uses a combination of back-scatter electrons (BSE) and energy dispersive spectrum (EDS) x-rays to identify, quantify and characterise materials. Filters are examined at a 0.5 µm resolution.

4.0 Monitoring strategy 4.1 Monitoring programme 2005 Programmes at each site ran for approximately two weeks during the period June – August 2005 (Appendix 1, Table 1). During each two-week period, four filters were exposed consecutively, each for 72 hours. 4.2 Monitoring programme 2006 In order to improve the robustness of the statistical analysis it was decided that a greater number of samples was required; hence for 2006 the monitoring period at each site was extended. Each programme ran for approximately four weeks from June – November 2006 (Appendix 1, Table 2). During each four-week period at each site, eight filters were exposed consecutively, each for a period of 72 hours. In the final monitoring programme an additional filter was exposed at each site. Due to equipment failure insufficient samples were collected at the Porthtowan site during the 2006 programme and to rectify this deficiency it was decided that instead of monitoring at Trelissick (background site) the site at Porthtowan would be revisited for a second monitoring period. 4.3 Filter analysis - 2005 In 2005 43 (out of 48) valid filters were retrieved and at the end of the monitoring period each filter was divided into two halves. One half of each filter was retained and the other halves were sent together in bulk to Royal Holloway for ICP-MS analysis. ICP-MS was used to quantify the arsenic collected on each filter. Nine of the retained filter halves (one per site) with the highest arsenic loading, according to the ICP-MS results, were sent to Camborne School of Mines (CSM) for QEMSCAN™ analysis. The remaining filters were sent to the University of Plymouth (UoP) for further ICP-MS and SEM analysis to corroborate the initial results. Due to operational error at UoP the ICP-MS results were not considered robust and were discounted. AQU: Arsenic in ambient air Final report December 2006

13

4.4 Filter analysis 2006 In 2006, 81 (out of 99) valid filters were retrieved. Filters from each of the monitoring programmes were retained until the whole survey was complete. Each filter was then divided into quarters and one quarter of each filter was sent to Harwell Scientifics (Didcot) and one quarter to Royal Holloway (London) for ICP-MS. The remaining half of each filter was retained by the AQU in case further analysis of specific samples should be required. As results from Harwell Scientifics and Royal Holloway for each filter were in agreement (Figure 4) it was decided that further analysis of the retained samples was not required. 4.5 Monitoring problems 2005 During the first year’s monitoring some problems were encountered, i.e. accidental power supply interruption and equipment failure. The Partisol (H1) at the control site (Stithians) experienced failure of the flow controller during Programme 3. This resulted in incomplete monitoring periods with the ‘valid’ pumping times falling short of the programmed 72 hours. Monitoring continued at the other sites (Falmouth and Carharrack) until Programme 3 was completed. During this programme, Casella service engineers failed to repair the unit on site. Unit H1 was returned to the office for testing and replaced with unit H2 to investigate the source of the problem. A new service team (Air Monitors Ltd) were contracted and Partisol H1 was sent to them for repair, thus delaying the commencement of Programme 4. 4.6

Monitoring problems 2006

During the 2006 programme problems due to equipment failure were also encountered. The unusually hot weather caused a software malfunction in the Partisol which was sited at Porthtowan. Repairs were made and the Partisol was relocated at the site for a final monitoring programme, instead of at Trelissick, as it was felt to be more important to obtain a valid set of data at Porthtowan (as it is an area where mining activity was intense and where mine spoil heaps are prolific). One filter at Stithians was lost (24/07/06) due to equipment failure. The fault was rectified by the servicing company (Air Monitors Ltd) and the programme was resumed the following week.

5.0 Results 5.1 2005 results 5.1.1

ICP-MS 2005

ICP-MS filter analysis results from Royal Holloway (University of London) produced arsenic concentration values in the range 0.08 – 2.78 ng m-3 (Appendix 2). The lowest values from the analysis are around the method’s limits of detection and the highest values agree with annual mean concentrations of arsenic in the UK rural environment (1 – 4 ng m-3) (Maggs 2000). Most results are below this range and all are well below the EU Air Quality guidelines of (6 ng m-3) (EC 2004). Figure 3 shows the site average concentrations of airborne arsenic over the whole 2005 monitoring period. Results from monitoring sites at Bissoe, Carharrack (Samsfield) and Camborne (Tuckingmill) show average site concentrations greater than at the control site (Stithians), the rural background sites (Constantine, Crowan and Trelissick) and the urban background site (Falmouth).


14

Average airborne arsenic per site

0.8

0.88

0.72

0.6

0.3 0.2

0.31

0.4

0.32

0.44

0.5

0.1

0.22

0.57

As (ng m-3)

0.7

0.34

0.9

0.95

1

Tr el is si ck

an Cr ow

h Fa lm ou t

St ith ia ns Po rth To w an Co ns ta nt in e

Ca rh ar ra ck Tu ck in gm i ll

Bi ss oe

0

Figure 3: Site average of airborne arsenic concentrations – Year 1 (2005) 5.1.2

QEMSCAN™ 2005

QEMSCAN™ results revealed only one filter (Carharrack (filter 30)) with three arsenicbearing particles (out of > 91,000 particles analysed) (two arsenopyrite and one arsenopyrite/Fe-oxides mixture). This is probably due to the heterogeneous distribution of particulate arsenic on the filter media (Almeida, et al. 2003). Also, QEMSCAN™ analysis detects particles > 0.5 µm whereas airborne arsenic is concentrated in fine particulates < 1.0 µm (EU 2001). 5.2 2006 results 5.2.1

ICP-MS 2006

The ICP-MS results obtained from the two laboratories (Harwell Scientifics and Royal Holloway), although not statistically related (Section 5.3) (Figure 4), were sufficiently similar to confer confidence in the analysis techniques used (see section 5.3). By indication this also gave validity to the 2005 ICP-MS results obtained from Royal Holloway. The range of arsenic concentrations observed in 2006 was greater than in the previous year (Royal Holloway (0.10 – 5.47 ng m-3, X ¯ = 1.03) with some monitored values higher than annual mean concentrations of arsenic in the UK rural environment (1 – 4 ng m-3) (Maggs 2000), and just below the EU Air Quality guidelines of 6 ng m-3 (EC 2004). Appendix 3 details the average arsenic concentrations per filter from the two laboratories. The highest concentrations were recorded at Bissoe and Carharrack (Samsfield) (5.2086 ng m-3 and 4.2841 ng m-3 respectively). Concentrations > 4 ng m-3 were also recorded at Camborne and Falmouth for one filter only per site (filters 3 and 5 respectively).


15

Scatterplot of Harwell vs Holloway 6 5

Harwell

4 3 2 1 0

0

1

2

3 Holloway

4

5

6

Figure 4: Scatterplot of ICP-MS results from Harwell and Royal Holloway with 1:1 reference line Figure 5 shows the site average concentrations of airborne arsenic for 2005 and 2006 (average results from Royal Holloway and Harwell Scientifics). At most monitored sites average airborne arsenic concentrations were higher in 2006 than 2005 by a factor of 2 – 3. By contrast, average concentrations for 2006 of arsenic at the control site (Stithians) and Constantine were slightly lower than those observed in 2005. No values are available for Trelissick for 2006 as the monitoring equipment was only deployed at this site during 2005. Average airborne arsenic per site 2.94

3 2.65

2005

2.5

1.68

-3

Figure 5: Site averages of airborne arsenic concentrations 2005 and 2006 AQU: Arsenic in ambient air Final report December 2006

16

0.22

Tr el is sic k

C

ro w an

0.32

Fa lm ou th

0.34 0.27

C on st an tin e

To wa n

Po rt h

St ith ia ns

ill Tu ck in gm

rra ck C ar ha

Bi ss oe

0

0.31 0.40

0.84

0.5

0.44

0.72

0.88

0.95

1

1.15

1.5

0.57 0.45

As (ng m )

2

2006

5.3 Statistical analysis In 2005, the number of valid filters obtained from each site was insufficient to perform statistical analysis. This was compounded by a broad variation between concentrations ¯ = 0.57, n = 11). To ensure recorded at each site (control range 0.80 – 2.78 ng m-3, X statistical robustness the sampling strategy in 2006 was adjusted and the sample numbers increased. The 2006 filter analysis results obtained from the two laboratories (Harwell Scientifics and Royal Holloway) indicated a high degree of correlation (r = 0.94) (Figure 4). However, a paired T-test revealed that the results were not statistically similar at the 95% confidence level and analysis of the differences showed a broad distribution of outliers about the mean. Given the high degree of correlation and the inherent heterogeneity of filter media capture methods (Almeida, et al. 2003) the results from both laboratories were, however, considered valid and a mean concentration was calculated per filter. To carry out statistical analysis of the 2006 filter analysis results the mean concentrations of arsenic per filter were normalised using their natural log. One-way analysis of variance (ANOVA) by site was performed on the normalised data and the means compared with the control site mean using Dunnett’s family error rate (5). The monitoring sites Bissoe, Camborne (Tuckingmill), Falmouth and Carharrack (Samsfield) showed statistically significant differences to the control site (Stithians) data. Statistical analysis was also performed on the logged arsenic concentrations against each of the monitoring units, the monitoring programmes, PM10 (recorded by the Osiris monitor) and meteorological data (total rainfall (mm) and average wind speed (knots)) but no significant trends were observed in relation to concentrations of airborne particulate arsenic. Further geographical investigation was carried out to determine whether the proximity of past mining sites had any influence on concentrations of airborne arsenic. The distance to all mining sites was calculated in ArcView GIS using the Spatial Analyst tool Euclidean Distance. The harmonic mean of each monitoring site’s distance to all mine sites in the EA database was calculated to give greater weight to mine sites within a closer proximity. No statistically significant correlation was shown to exist between the 2006 monitored airborne arsenic values and the average distance to mining sites. Density of mining sites was calculated in ArcView GIS using the Spatial Analyst tool Point Density (Table 2). Density (mines km-2) was calculated using the mine sites within a 5 km2 neighbourhood at a 10 m resolution (Figure 6). No statistically significant correlation was shown to exist between the 2006 monitored airborne arsenic values and the mining density. Table 2: Mining site density of monitoring sites Monitoring site Mining site density Bissoe 0.96 Camborne 2.24 Carharrack 2.48 Constantine 0.28 Crowan 0.92 Falmouth 0.16 Porthtowan 1.40 Stithians 0.40 Trelissick 0.04


17

Figure 6: Density of mining sites (mines km-2)

6.0 Discussion 6.1 2005 programme The first year’s (2005) ICP-MS results showed airborne arsenic concentrations in the range 0.08 – 2.78 ng m-3. Average site concentrations were all < 1 ng m-3 and the highest concentrations of airborne arsenic were found at Bissoe, Carharrack (Samsfield) and Camborne (Tuckingmill)). Although the data is not statistically robust, these sites are close to mine sites and exposed mining spoil (Figure 6). The highest individual filter arsenic concentrations were recorded at Stithians (2.78 ng m-3) and Carharrack (2.32 ng m-3) during the same exposure period in Programme 3 (2005). These values are disproportionately higher than other concentrations from these sites, thus negatively skewing these sites’ average concentrations. Although no statistical test could be carried out on the small dataset the elevated concentrations of arsenic monitored during this exposure period do not appear to be reflected in PM10 or meteorological data. Technical faults with the Partisol monitor located at Stithians were identified during this exposure period but, given the elevated arsenic concentrations simultaneously recorded at Carharrack, this is not thought to have influenced the results obtained. QEMSCAN analysis confirmed the presence of arsenic particles on the filter exposed at Carharrack during this period although no quantitative data was obtainable from this technique. However, only three arsenic-associated particles were found at the Carharrack site. QEMSCAN™ detects particles > 0.5 µm whereas airborne arsenic is concentrated in fine and ultrafine particulates (< 1.0 µm) (EU 2001; Culbard and Johnson 1984). A number of monitored arsenic-related particles may therefore have been below the limits of detection of AQU: Arsenic in ambient air Final report December 2006

18

the QEMSCAN instrumentation. The implications for human health of exposure to fine arsenic-contaminated particles are particularly significant as particulates < 1.0 µm are able to cross the gas-blood interchange and enter the bloodstream. Due to the significant cost of analysis and the unsuitability of the technique for identifying particulates in the fine–ultrafine fraction, QEMSCAN analysis was not repeated in 2006. Quantitative corroboration of the ICP-MS results obtained from Royal Holloway (University of London) for 2005 was sought from University of Plymouth; however, operational error at the University of Plymouth laboratory negated the comparison. 6.2 2006 programme Prior to the 2006 monitoring survey a set of six trial filters (exposed at Rosewarne, Camborne) were run to test the monitoring equipment. These filters were quartered and sent for ICP-MS analysis to Royal Holloway and Harwell Scientifics. Although the sample size was too small to perform statistical correlation and the concentrations recorded were at the limits of detection of the technique, the results obtained from the two laboratories were within a similar range, conferring confidence in the analysis techniques employed by both analysts and inferring confidence in the 2005 results obtained from Royal Holloway. Confidence in the two laboratories used (Royal Holloway and Harwell Scientifics) was confirmed by comparison of the 2006 ICP-MS results reported. A 94% correlation was observed between Royal Holloway and Harwell Scientifics despite statistically significant differences between filter results identified by a paired T-test. Given the very high correlation a degree of error was accepted as it is generally accepted that particulate matter is not homogeneously distributed across the filter media (Almeida, et al. 2003). An average of the ICP-MS results from the two laboratories for each filter provided arsenic concentrations within the range of 0.01 – 5.47 ng m-3. Monitored data at Bissoe in 2006 gave the highest concentrations of airborne arsenic with two filters recording concentrations > 5 ng m-3. Although these individual filter concentrations are not directly comparable with legislative annual mean limit values (6 ng m-3) (EU 2001) the consistently higher values at Bissoe over both years’ monitoring programmes would warrant further monitoring at this site. It is worthy of note that elevated concentrations recorded at Bissoe were not reflected in concentrations at other simultaneously monitored sites in Programme 4 (2006). Relatively high arsenic concentrations for Programme 2 (2006) were also recorded at Carharrack (Samsfield) on several filters. Elevated concentrations (> 4 ng m-3) were observed at Falmouth for one filter exposure. The Falmouth filter had particulate arsenic concentrations > 3 times higher than any other filter from this site. Both portions of the filter analysed have highly consistent readings (difference 0.064 ng m-3). Explanations of this unusually high concentration may be due to a source of arsenic relating to localised activities. 6.3 Comparison of 2005 and 2006 data Average site concentrations recorded in 2006 show higher values than in 2005 at most sites, however lower concentrations were recorded at Stithians (control site) and Constantine, and Crowan’s average results differ only marginally over the two years (difference + 0.09 ng m-3). At present it is not clear why some sites have recorded higher concentrations of airborne arsenic in 2006. The greatest differences were observed at Bissoe, Carharrack and Camborne. Each of these sites is close to mining sites and areas of exposed mining spoil. Using mine sites as a proxy for exposure to arsenic-contaminated land, no statistical correlation was identified between airborne arsenic concentrations and average distance to all mine sites (within the EA database) or mining density. Specific contaminated land data was not available at this time, however further analysis is recommended to determine whether there is a statistical AQU: Arsenic in ambient air Final report December 2006

19

relationship between airborne arsenic concentrations and proximity to exposed contaminated soils. Further monitoring at Bissoe, Carharrack and Camborne is also recommended to ascertain whether these higher concentrations in 2006 are resultant from resuspension of contaminated dust particles.

7.0 Conclusion The purpose of this investigation into concentrations of airborne arsenic in Cornwall was to establish a baseline of data against which further monitoring of particulate arsenic may be compared. This preliminary research has indicated that ambient concentrations of airborne particulate arsenic in Cornwall are below WHO guidelines and EU legislation limit values (6 ng m-3). An elevated concentration of arsenic recorded at some monitored sites has raised questions regarding the disparity of potential human exposure in certain areas in Cornwall. The geographical pattern of airborne arsenic concentrations requires further analysis against contaminated land and geological datasets to determine whether the proximity of these factors has any statistical influence.

8.0 Further research As mentioned above, further analysis of the monitored results is recommended to investigate the relationship between ambient airborne particulate arsenic and contaminated land and geological data. This could be combined with additional monitoring at sites in close proximity to past mining activity and exposed mining spoil, e.g. Bissoe, Carharrack and Camborne. It may also be worthwhile monitoring at other sites in Cornwall and the UK to investigate whether the results obtained from this investigation are applicable elsewhere. Long-term monitoring at these sites would provide a more statistically robust dataset and may indicate the presence of temporal patterns in the airborne arsenic concentrations. The brevity of the monitoring undertaken in this project means that the results are only indicative and cannot be accurately compared with EU limit values relating to annual mean concentrations of airborne particulate arsenic. The EU 4th Daughter Directive 2004/107/EC is due to be adopted into UK legislation in 2007 and long-tem monitoring may be required to comply with revised National Air Quality Strategy objectives. In addition to long-term monitoring it would be beneficial to monitor all sites simultaneously to ensure inter-site comparability. The scope of this project was severely limited by access to monitoring equipment; however, subject to funding criteria, additional equipment could be hired on a long-term lease if further research was undertaken in this area. Another area of potential research is an investigation into the ingress of airborne particulate matter to the indoor environment.

9.0 Acknowledgements The Air Quality Unit at Cornwall College would like to acknowledge the Environment Agency and the European Social Fund for their funding contributions to this investigation. We would also like to thank the site hosts who housed the monitoring equipment and the Universities of Exeter (Camborne School of Mines), Plymouth (Electron Microscopy Centre) and London (Royal Holloway) and Harwell Scientifics for filter analysis. This programme is part funded by the European Union through the European Social Fund – helping to develop employment by promoting employability, business spirit and equal opportunities and investing in human resources.


20

10.0 Bibliography Abrahams PW and Thornton I (1987) Distribution and extent of land contamination by arsenic and associated metals in mining regions of southwest England. Journal of Geochemical Exploration 82, 2 Almeida, SM, Freitas, MC, Reis MA, Pio CA (2003) Quality assessment on airborne particulate matter of k0-INAA. Journal of Radioanalytical and Nuclear Chemistry 257, 3:609-613 Archer FC and Hodgson IH (1987) Total extractable trace element contents in soils in England and Wales. Soil Science 38, 421–431 CAQF (Cornwall Air Quality Forum) (2003) Continuous monitoring of airborne particulate matter (PM10) at United Downs, Cornwall, 2003. http://www.cornwall-airquality.org.uk/pdf/udowns.pdf [Accessed 3/11/06] CEC (Commission of the European Communities) (2003) Directive of the European Parliament and of the Council relating to arsenic, cadmium, mercury, nickel and polycyclic aromatic hydrocarbons in ambient air. http://www.europarl.eu.int/meetdocs/committees/envi/20040120/509184en.pdf [Accessed 12/3/04] Cornwall County Council (2004) Metalliferous Mineral Extraction. http://www.cornwall.gov.uk/Environment/minplan/mlp11.htm [Accessed 21/01/05] Culbard EG and Johnson LR (1984) Elevated arsenic concentrations in house dusts located in a mineralised area of southwest England: Implication for human health. Proceedings of the University of Missouri 18th Annual Conference “Trace Substances in Environmental Health XVIII”, Ed. Delbert D Hemphill, June 4–7 1984, University of Missouri, Columbia, USA, 311-319 DEFRA (Department for Environment, Food and Rural Affairs) and the EA (Environment Agency) (2002) Contaminants in Soil: Collation of Toxicological Data and Intake Values for Humans. Arsenic. http://www.environment-agency.gov.uk/ commondata/acrobat/tox1_arsenic_675423.pdf [Accessed 21/07/2005] DEFRA (2004) Copper Chrome Arsenic Regulatory Impact Assessment. http://www.defra.gov.uk/corporate/regulat/ria/2004/cca.pdf [Accessed 27/07/2005] EA (Environment Agency) (2002) Soil Guideline Values for Arsenic Contamination. http://www.environment-agency.gov.uk/commondata/acrobat/tox1_arsenic_675423.pdf [Accessed 21/07/2005] EC (European Commission) (2004) 4th Air Quality Daughter Directive. Council Directive 2004/107/EC EU (European Union) (2001) Office for Official Publications of the European Communities. Ambient air pollution by As, Cd and Ni compounds. Position Paper. Luxembourg: ISBN 92-8942054-5 Foster R (2004) Dust clouds gather over landfill extensions work. The West Briton. 1st July, p. 29 ISO 7708:1995 (1995) Air quality -- Particle size fraction definitions for health-related sampling. Technical Report ISO/TC 146/SC 2 AQU: Arsenic in ambient air Final report December 2006

21

IARC (The International Agency for Research on Cancer) (1998) Kavanagh P, Farago ME, Thornton I, Goessler W, Kuehnelt D, Schlagenhaufen C and Irgolic KJ (1998) Urinary arsenic species in Devon and Cornwall residents, UK. A pilot study. The Analyst. 123, 27–29 Maggs R (2000) A Review of Arsenic in Ambient Air in the UK. . Department of the Environment, Transport and the Regions, Scottish Executive, The National Assembly for Wales. Stationery Office, London. Moorcroft S (1997) £500,000 study signals new DOE interest in fine particles. Air Health Strategy. May 1997, p1 Thornton I and Farago M (1997) The geochemistry of arsenic. From Abernathy CO, Calderon RL and Chappel WR, Arsenic, Exposure and health effects, Chapman and Hall, Chapter 1, p1 Thornton I, Abrahams PW, Culbard E, Rother JAP, Olsen BH (1986) The interaction between geochemical and pollutant metal sources in the environment: implications for the community. Journal of Geochemical Exploration 82, 2 WHO (World Health Organization) (1997) WHO Air Quality Guidelines. Arsenic (Draft). World Health Organization, Geneva. WHO (World Health Organisation) (2000) Air Quality Guidelines 2nd Edition. http://www.euro.who.int/document/e71922.pdf [Accessed 27/07/2005] WHO (2003) Arsenic and Arsenic Compounds. http://www.inchem.org/documents/ehc/ehc/ehc224.htm [Accessed 12/3/04] World Heritage Cornwall (2006) Bissoe Arsenic Works, Cornwall. http://www.worldheritagecornwall.com/mines/bissoe-arsenic-works.htm [Accessed 18/12/06] Xu J and Thorton I (1985) Arsenic in garden soils and vegetable crops in Cornwall, England: implications for human health. Environmental Geochemistry and Health 7, 131–133


22