Annals of Occupational Hygiene Advance Access originally published online on September 19, 2006
Annals of Occupational Hygiene 2006 50(7):717-729; doi:10.1093/annhyg/mel064
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Characteristics of Pesticide Pyrotechnic Smoke Devices
1 Health and Safety Laboratory, Harpur Hill Buxton, Derbyshire, SK17 9JN, UK
2 Health and Safety Executive, Stanley Precinct Bootle, Merseyside, L20 3QZ, UK
*Author to whom correspondence should be addressed. E-mail: martin.roff{at}hsl.gov.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Pesticide smoke generating products are widely used by amateurs and professionals but there is little published information available about their burn and deposition characteristics to enable the risks associated with using these devices to be assessed. This paper investigates their burn characteristics, deposition patterns, pesticide air concentrations and potential exposure to operators. Thirteen firings were carried out in different spaces with different ventilation conditions. Three types of devices were investigated: dicloran, permethrin and red dye. Pesticide air concentrations increased after firing, reaching a maximum determined by the room volume in
10 min and decreasing exponentially as a result of ventilation and deposition. Ejected pesticide was present in the aerosol phase but there were only occasional traces of vapour. Settlement of pesticide was affected by surface orientation, height, sampling material and the pesticide-to-space volume ratio. The manufacturer's recommended treatment period for dicloran of 4 h followed by half an hour of ventilation may be insufficient to reduce pesticide to safe levels for re-entry under very calm conditions.
Keywords: pesticide • pyrotechnic • ventilation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Under the Control of Pesticides Regulations (COPR) all pesticides must be approved before they can be advertised, sold, supplied, stored or used in the UK. Responsibility for approval, assessment and regulation is split between two agencies: the Pesticides Safety Directorate of the Department for Environment, Food and Rural Affairs (DEFRA; agricultural pesticides), and the Biocides and Pesticides Assessment Unit Pesticides Registration Section of the Health and Safety Executive (HSE; non-agricultural pesticides). DEFRA and HSE require information to assess the risks associated with the use of these devices in accordance with their duties under the EC Biocides Directive (EC, 1998).
Pesticide pyrotechnic smoke generating products contain roughly equal proportions of pesticide and pyrotechnic. When ignited they produce thick smoke, dispersing the pesticide by turbulent diffusion and convection for rapid insect knockdown in the air space and for surface pesticide residue. They are used to apply insecticides and fungicides in agricultural and other environments, often where spray application is difficult or impractical e.g. in large buildings and in roof or floor voids. They are also used to treat greenhouses, ships' holds and enclosed food storage spaces such as grain stores.
There is little published information on the burning, dispersion and settlement characteristics of these products. HSE commissioned this study to characterize the behaviour of pesticide smoke generators under controlled and real application conditions so that risks during use could be better understood.
The objectives of this study were to determine burn characteristics, smoke concentrations, dispersion characteristics, smoke deposition patterns and operator exposure on re-entry after the supplier's recommended exclusion period.
The study was carried out in three phases: Phase 1—five repeated releases inside an 84 m3 test cell; Phase 2—a single release in a 1300 m3 space, and Phase 3—five releases in 213 and 293 m3 spaces. Ventilation measurements were made prior to the tests. The devices were fired centrally within the test spaces where sampling and monitoring instruments were positioned. The test room was vacated and the doors were closed or sealed. After the exclusion period of 4 h had elapsed, the test spaces were re-entered briefly by operators wearing suitable PPE to ventilate the space until smoke had completely cleared. Samples and instruments were then collected. To further investigate the initial burn characteristics of the devices, some additional, short tests (Phase 1a) were carried out in the same test cell as the Phase 1 tests but without the 4 h exclusion.
Phase 1: 84 m3 test cell
An enclosed 84 m3 test cell 3.8 m high, was lined with a timber batten frame and covered with polythene sheeting (Fig. 1). Removable polythene flaps allowed forced ventilation through extract fans. An array of 16 vertical support poles (3 m high) was set up inside the chamber for the deposition sampling devices and instruments. The device was fired electronically from outside the cell. Phase 1 tests investigated scaled down versions (rated at 62 m3) of commercial permethrin, dicloran and red dye products which under-dosed the test space by 24% when compared with the recommended doses of 90 mg m–3 and 20 mg m–3 for dicloran and permethrin, respectively. The red smoke generators were investigated to see if they were suitable as surrogates for pesticide smoke generators for future simulations.
|
Phase 2: 1300 m3 real space
An isolated storage building was used as a typical application space. It was of irregular shape, being 1300 m3 volume and 8.3 m high in the central area, with two 4 m high offshoot areas on either side (Fig. 2). Four full-sized (500 m3) commercial dicloran devices were ignited manually, which over-dosed the space by 54% compared to manufacturer's recommendations. The room was vacated and the doors closed but not sealed. Four support poles (A–D) were used for deposition samplers. Sampling instruments were mounted on pole D in the centre of the room.
|
Phase 3: 213 and 293 m3 real spaces
Two adjacent rooms, both 2.96 m high, were used for the Phase 3 tests, connected by double doors (Fig. 3). The larger room (213 m3) had a boiler room in one corner, two opening windows, a side door into the interior of the building, one pair of double doors opening to the outside and one pair opening into the smaller room (79 m3). Three support poles (A–C) were used for deposition samplers and sampling instruments. The floor was covered with paper. Equipment stored in the room was covered with polythene sheeting. Rooms were both under-dosed and over-dosed in Phase 3 tests using devices rated at 250 and 125 m3.
|
Test 3 was notable for the exceptionally calm conditions prevailing on the day.
After the tests, all materials used to line the test spaces were rolled up with exposed surfaces inwards and bagged and labelled for safe disposal. Remaining surfaces were de-contaminated by vacuuming using HEPA filters. Wipe samples were collected from horizontal surfaces before and after vacuuming, which showed that >90% of the contamination had been removed. In Phase 2, some surfaces were not de-contaminated. These were sampled 11 months later to determine the residence time of pesticide residues on horizontal surfaces.
|
EXPERIMENTAL EQUIPMENT AND MEASUREMENT METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Air change rate measurements
Air change rates (ACRs) were measured in the test spaces prior to the tests using the method of MDHS 73 (HSE, 1992). Sulphur hexafluoride tracer gas was released for a short period of time and mixed well in the test space using fans. A Miran 1A infrared gas analyzer continuously monitored the gas concentration C over time t. The gradient of the log/linear plot gave the air change rate
in units of per hour (h–1), as the exponential decay time constant from the equation Ct = C0. e–t.
Burn characteristics (Phases 1 and 1a)
The rate of loss of material during the test was measured by firing the device mounted on an electronic balance (Sartorius LC6200S), suitably protected against heat, measuring at 1 s intervals via an RS 232 link on the floor in the centre of the test cell. A second electronic balance was used to weigh each device before and after the test to determine the total weight loss more accurately.
The velocity and temperature of the plume were measured at various heights directly above the device using hot wire anemometers and 0.25 mm diameter thermocouples as shown in Fig. 4. The anemometers were calibrated in a wind tunnel before and after each test whenever possible, to ensure that their responses had not been affected by particles adhering to the hot wires. Other thermocouples placed near the corner of the cell recorded air temperatures at heights of 1, 2 and 3 m above the floor. Readings were logged every second for 5–10 min after ignition.
|
Velocity measurements were also made at 3.7 m (10 cm below the ceiling level of the test cell) using ultrasonic anemometers (Gill Instruments ‘Windmaster’) to measure the time of flight of pulsed sound waves across a 14.6 cm path, from which the three orthogonal air velocity components were calculated. The data were logged on computer every second for 15–20 min after ignition.
The Phase 1 measurements described above were made during the Phase 1 and Phase 1a (short) tests. The Phase 1, 2 and 3 tests continued with 4-h sampling for air and surface deposition concentrations as described below.
Air concentration measurements
Air samples were collected according to MDHS 94 (HSE, 1999) using between 3 and 5 programmable pumps sampling at 0.5 l min–1, at timed intervals throughout the duration of each test. Each pump head contained a 25 mm glass fibre (GF/A) filter to collect pesticide aerosol, backed up by a glass tube containing Tenax absorbent to collect pesticide vapour. Samplers were bunched and taped to one support pole at 0.5 m above the floor.
The aerosol concentration in the air, which included aerosolized smoke and pesticide particulates and droplets but excluded vapour, was logged every 5 s in units of milligrams per cubic metre (mg m–3) on two battery-operated DustTrak light-scattering dust monitors sampling continuously at 2 l min–1 mounted on support poles at the top or at the mid-pole height. The data were downloaded to computer after each test. Two Sierra multistage impactors mounted on support poles at the top or at the mid-pole height, sampled the air at 2 l min–1 via pumps in Phases 1 and 2. Each stage sampled a particular size fraction of the aerosol. Quantitative analysis was performed on rinsings from each stage. The four instruments combined would have pumped 2 m3 in a 4.5 h test which should not have caused significant air movements or disturbances.
Deposition sampling devices
Sixteen 3-m high vertical support poles were constructed from square tubular steel for mounting the deposition samplers and air monitoring equipment. In Phase 1, all 16 were positioned in the test cell as shown in Fig. 1. Balsa-wood blocks were used to provide surfaces for measuring deposition and these were mounted at three heights (15, 150 and 300 cm) on one side of the vertical support poles.
The corners of the blocks were trimmed at 45° to provide five surface orientations at 0°, 45°, 90°, 135° and 180°, onto each of which pairs of 10 mm diameter deposition samplers, or single 25 mm diameter samplers could be fixed. The 10 mm samplers were drawing pin heads representing smooth surfaces, paired with glass fibre GF/A filters representing rough surfaces similar to blotting paper affixed with Blu-Tac. Both surfaces were selected for their low electrostatic charge. Additional 25 mm diameter GF/A samplers were placed at 0° only, representing larger rough surfaces. These were protected from transfer from the contaminated block by fresh 35 mm diameter GF/A samplers placed underneath.
In the first two tests of Phase 1, 10 mm rough and smooth samplers were fixed at all five orientations. Deposition was found to be negligible at 135° and 180°, so these surfaces were not sampled subsequently. There were no differences seen between the rough and smooth surface types in the first three tests of Phase 1, so only 25 mm diameter GF/A samplers were used subsequently in Phases 1, 2 or 3.
In Phase 2, four support poles were used, and in Phase 3 only three. In Phase 3, the support poles were reduced in height to accommodate the lower ceiling, so the highest block was at 257 cm. In Phase 3, only the 0° surfaces of the blocks were used for sampling but additional polythene and GF/A samplers were mounted vertically on the walls to compare the effects of different materials on deposition (Fig. 3).
Personal sampling on re-entry (Phase 3 only)
During re-entry to open doors and windows for ventilation, the residual airborne dust was observed with a Tyndall light beam (MDHS 82; HSE, 1997). The operator wore a filtering face piece half-mask (FFP3) and a clean Tyvek® oversuit and a 7 cm diameter GF/A filter paper was pinned at the collarbone to act as a passive surface sampler. A personal pumped sampler head, sampling at 0.5 l min–1 and containing a 25 mm GF/A filter, was fitted at shoulder level and worn for the few minutes of re-entry, backed up by a Tenax tube. All the samples were removed immediately after the short re-entry period, bagged and stored frozen prior to analysis.
Video records
In Phase 1, one video camera was used to record the firing of the device and the direction of the smoke jet, and a second was used to record the ceiling to monitor the spread of the smoke plume under floodlights inside the test cell. The lights were only used for brief intervals so as not to introduce heat and disturb the air flow. In Phases 2 and 3, cameras were used to monitor the spread of the plumes continuously in natural light. The videos showed that the smoke spread over the ceiling at first, leaving the lower half clear for up to an hour in still conditions, and then gradually descended to fill the entire chamber as time progressed.
|
ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Cyclohexane containing 2 µg ml–1 dieldrin as an internal standard was used as an extraction solvent for dicloran and permethrin. Tetrahydrofuran (without an internal standard) was used as an extraction solvent for red dye. Appropriate volumes were pre-measured into labelled vials prior to collection of the various sampling devices on-site following each test. Impactors were transported intact back to the laboratory, stored frozen and each stage samples were extracted just prior to analysis.
The pesticide extracts were analysed by gas chromatography with electron capture detection (GC-ECD) employing a PAS 1701 column (14% cyanopropylphenylmethylpolysiloxane). The red dye extracts were analysed by high performance liquid chromatography (HPLC) using an ODS 2 column, a mobile phase of tetrahydrofuran: water: glacial acetic acid at ratios of 16:24:1, and an ultra-violet (UV) absorbance detector set at 254 nm. The extracts were further diluted and re-analysed when high concentrations of analyte were initially found.
The remaining pesticide content of the expired devices were determined after firing by scraping residues from the pyrotechnic casings and dissolving in 30 ml of cyclohexane or 100 ml of ethyl acetate. Further dilutions were made to enable quantitative analysis. These results were compared with the original composition figures, in order to calculate the mass of pesticide ejected or burned.
|
STATISTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Statistical analysis of the surface deposition results was by General Linear Modelling using Systat®. The block locations, the surface orientations and the surface types were used as categorical independent variables. In Phase 1, effects of surface orientation and sampler type were analysed within-blocks, and the effects of block location (expressed as row, column and height), distance and facing relative to the device were analysed between blocks. In Phase 2, the within-block analysis was similar to Phase 1, but the four poles A–D were not in a row/column array. In Phase 3, the test numbers (1–5) and the sampler heights (highest, middle and lowest) were used as categorical independent variables, and the three pole locations A–C were nested within-test.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Burn characteristics
Table 1 shows the burning times, weight losses, maximum emission rates, maximum temperatures and particle size distributions from Phases 1 and 1a combined. Device firing characteristics were not captured on video in Phase 2. In Phase 3, burning times were measured from the video footage (i.e. not determined gravimetrically). Temperature variations in Table 1 may be explained in part by the video evidence that showed that occasionally and unpredictably, the plumes veered away from the vertical, so the thermocouples were not always in their centres. Permethrin devices are designed to burn at lower temperatures (233°C) than dicloran or red dye (283 and 392°C), due to the physical state of the pesticide.
|
Large differences were seen for the maximum emission rates from dicloran devices. In two tests the generators burned very slowly and erratically, whereas in another the same type of device had a high emission rate. The red smoke generators burned more energetically than the other devices, producing a measurable down-thrust on the balance. Measured plume velocities above the fired pyrotechnic device were 0.5–1 m s–1.
Particle size distributions
Most of the particles from both full scale and smaller versions of the devices were contained in only a few of the stages of the impactors, corresponding to aerodynamic diameters of between 0.5 and 3 µm. The particle sizes were small, with 99% <4.25 µm and thus classed as respirable. Mass median diameters (d50) were 1.3 µm for permethrin and between 2 and 2.5 µm for dicloran and red dye (Table 2). Typical size distributions for each type of device are shown in Fig. 5. This shows that red dye is a good surrogate for dicloran for settlement because their smokes have similar particle size distributions.
|
|
Aerosol and pesticide concentrations
The test conditions in each phase and a summary of the devices used, the test space volume, the ventilation characteristics and average deposition results for upward facing horizontal samplers (0°) are given in Table 1. Also, ventilation characteristics for some typical application spaces are given for comparison.
In all tests, pesticide concentrations increased for 10 min as the smoke dispersed following ignition, then decreased exponentially with time as a result of natural ventilation and deposition. A typical profile (Fig. 6) also shows that additional ventilation quickly reduced the aerosol below measurable levels after the 4-h treatment period. Pesticide or red dye accounted for about half of the aerosol mass (Fig. 6). Pesticides were only present in trace quantities in the vapour phase during a few of the tests.
|
The clearance of smoke from the air, defined here as the smoke clearance rate (SCR), is derived from the decay of aerosol concentrations in the same way that the ACRs were derived from decay of tracer gas concentrations. The ACRs and SCRs for all tests can be seen in Table 2, except for Phase 3 test 3 where DustTrak data was not recorded. In all of the unsealed and sealed tests, the SCRs were found to be slightly higher than the ACRs, which can be attributed to settlement. The ventilated test 1 of Phase 3 was an exception because the ACR was the same as the SCR (3.2 h–1), which indicated that settlement made no significant contribution to the SCR relative to the ventilation.
The maximum aerosol and pesticide concentrations at the fixed location on the support pole are shown in Fig. 7a and b, obtained from DustTraks and timed sampling pumps. Results are standardized by dividing the available pesticide dose in the device by the available room volume. The maximum concentrations depended on the amount of pesticide in the device per unit volume. Ventilation characteristics are described here as ‘sealed’ when the ACR = 0, ‘unsealed’ when there was slight natural ventilation 0 < ACR < 1, and ‘ventilated’ when natural ventilation from opened windows and seals etc. created an ACR > 1.
|
Red dye maximum aerosol concentrations were slightly higher than those for the pesticide devices (Fig. 7a). Ventilation had no effect on the maximum aerosol concentrations because the DustTrak maximum readings were in real time and measured the sharp peak during the initial few minutes as the plume gathered at the ceiling before ventilation took hold (Fig. 6). The effect of ventilation (ACR = 3.2 h–1) is apparent as a low result (
) in Fig. 7b where the maximum pesticide concentrations were averaged over 20 min of pumped sampling.
Average deposition
The average deposition over 4 h was found to depend on the available pesticide dose per unit floor area and the natural ventilation characteristics of the test space. Fig. 7c shows the average deposition for upward facing (0°) orientation samplers. In Phase 1 the test cell was under-dosed by 24% compared to supplier's recommendations. The average 0° deposition was 1.7 µg cm–2 in the ‘unsealed’ tests. In Phase 2, the ‘unsealed’ test space was over-dosed by 54% and the average 0° deposition was 13.5 µg cm–2: much more than twice the equivalent Phase 1 ‘unsealed’ test that might be expected from dosing considerations alone. However, the recommended dose is specified by room volume, which increases as the cube, whereas the floor area only increases as the square. Thus the available dose per unit area (and hence deposition) greatly increases with room volume.
Deposition patterns
A small sprinkling of larger particles of ash of 0.5 mm diameter, surrounded the device after firing. These were recovered from the paper in Phase 1 and analysed for pesticide content. Further afield, proximity to the device had no effect on deposition in Phase 1, nor did the direction of facing of the sampling blocks relative to the device firing position, i.e. whether facing towards or away from the device. Rough texture (GF/A) and smooth (metal) surfaces showed no difference in deposition. This confirms findings by Lai et al. (2002) for larger particles in a stirred chamber, albeit for a much rougher surface of 2 mm height.
In Phase 1 tests, mean pesticide deposit was between 1.7 and 3 µg.cm–2 on upward facing 0° samplers depending on the ventilation characteristics of the tests. The deposits on 45° surfaces were less than on 0° samplers by a factor close to cos 45° (0.7). Surfaces at 90°, 135° and 180° had collected negligible deposits. This pattern would be expected if vertical settlement under gravity in tranquil conditions was the dominant process, with only a minor contribution from air movement and electrostatic forces. Figure 8 shows the deposition at three heights and 0°, 45° and 90° orientations of all samplers in test 1 of Phase 1.
|
Deposition on the ceiling (180°) by impaction on GF/A filters vertically above the devices is shown in Table 1. There was much less deposition in the ventilated test because the smoke drifted away from the vertical during ascent and missed the samplers. Despite the direct impact during firing, the ceiling deposition was only 10—30% of the 0° samplers.
Deposition on different wall mounted materials was determined in Phase 3 tests. GF/A samplers of 7 cm diameter and 10 x 10 cm polythene sheet samplers were positioned vertically at head height as shown in Fig. 3. Table 1 and Fig. 9 show the deposition. GF/A samplers collected between 1 and 10% of that deposited on the 0° samplers, were consistent with the Phase 1 tests for 0° and 90° for GF/A. Polythene sheet collected 10 times more dicloran per unit area than the adjacent GF/A filter. Therefore deposition on vertical surfaces in calm conditions appears to be dominated by electrostatic forces.
|
In Phases 1 and 3, deposition varied with sampler height. Samplers positioned close to the ceiling (highest) generally collected more deposit than the middle and lowest samplers. The deposition on the lowest samplers was more erratic. In Phase 2, uniform deposition was found at all three sampler heights, but as the ceiling height was 8 m, and the highest sampler at 3 m was still relatively close to the floor, there was no measurable difference between the highest and lowest samplers.
As reported above, the mass median particle size of pyrotechnic smokes was 1.3–2.5 µm. Using Stoke's Law, gravitic settlement velocities of 1.3 and 2.5 µm particles are 0.006 and 0.02 cm s–1, respectively (Hinds, 1999). At the start of the 4-h test period, these particles were thrust to the ceiling, and would have fallen 86 and 290 cm, respectively due to gravity alone under tranquil settling conditions. Air movement was thermally induced when the devices fired, so smoke gradually mixed throughout the upper half of the test rooms from ceiling to floor over the first hour. It was noticeable that in Phase 1 ‘sealed’ tests, the settlement on the lowest samplers was proportionally much lower for the smaller permethrin particles than for larger dicloran or red smoke particles, and the vertical surfaces gathered little deposit, indicating that conditions were more characteristic of tranquil settling than stirred. The exponential decay of smoke (Fig. 6) suggests that stirred settling took place in the upper half of the chamber, but as the aerosol was polydisperse, a degree of decay would still be seen as larger particles drifted downward faster under tranquil settling. This is consistent with the video records.
Personal sampling devices on re-entry after 4 h (Phase 3 only)
In Phase 3, PPE and personal samplers were worn by the operator for 5–10 min on re-entry. The analyses are shown in Fig. 10.
|
Exposure to operators was in proportion to the aerosol measurements on re-entry after 4 h, although the 7 cm GF/A personal patch sampler on the very still day was a notable and unexplained exception. The Tyndall beam and the aerosol measurements after 4 h on the very still day (test 3) showed especially high concentrations of remaining aerosol. An additional Tyndall beam observation at 4.5 h showed that half an hour of ventilation with doors and windows wide open had had little effect. In similar conditions, unprotected operators would be exposed to pesticide on re-entry despite having followed the supplier's recommended treatment procedure. Therefore it is important that re-entry procedures should be included in manufacturer's instructions for use of the devices and that they also take ventilation and prevailing weather conditions into account.
Wipe samples from Phases 2 and 3 surfaces pre and post-cleanup
A high concentration of dicloran (4.8 µg cm–2) was collected on polythene sheeting still in place 11 months after the Phase 2 firings. This shows that dicloran has a long residence time on certain materials. In Phase 3, vacuuming reduced levels of dicloran on the windowsill from 5.4 to 0.6 µg cm–2.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Approximately 50% of the pesticide was consumed during the firing, so the composition of the device affects its burn characteristics. Measurement of the burn characteristics was hampered by the plumes veering erratically and the occasional side jet.
The generated smokes had consistent particle sizes and contained 50% pesticide and 50% ash. Air concentrations can thus be conveniently measured in real time using a light-scattering instrument. For the purposes of conducting future tests, devices containing less toxic red dye could be used instead of dicloran.
The residual air concentrations that were obtained under different ventilation conditions may be used to inform future guidance on timing and ventilation requirements prior to safe re-entry. Ventilation was an important factor in determining the clearance time within a treated space. Therefore the device instructions should recommend that operators check that aerosol has cleared (e.g. with Tyndall light beams) before the space is re-occupied even when manufacturer's recommended exclusion times have been adhered to. External weather conditions may also need to be taken into account.
Settlement of dicloran onto horizontal surfaces increased with the mass of pesticide available in the device(s), per unit of available floor area, and suggested a linear relationship. Maximum pesticide and aerosol concentrations were shown to suggest a similarly linear relationship to the mass of pesticide in device(s) as corrected for available volume in a treated space. Decay of the aerosol followed an exponential after an initial rapid fall, as found by Rowley and Crump (2005) for larger aerosol sprays. These data may be used by regulators to predict maximum acceptable air concentrations and surface deposits.
Large differences were seen for short-term deposition on different sampler materials, with plastics attracting up to 10 times the deposit of other materials.
The cleanup regime after all firings showed persistence of dicloran deposits. Vacuuming of horizontal surfaces reduced concentrations by a factor of 10. The residence time of pesticides on different surface materials may vary considerably. This could leave ‘hot spots’ of pesticide on re-entry with long residence times.
The slow settlement of the smoke particulate may have implications for areas with higher ventilation characteristics, where significant environmental releases may occur and also result in poorer efficacy. The deposition results indicate that pesticide smokes are unlikely to provide long term efficacy on anything other than upward facing surfaces. However, they may provide a knockdown effect during application.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The authors would like to thank the HSL Buxton workshops for their considerable efforts in preparing the test cell and the sampling support poles, and Octavius Hunt Ltd. for providing smoke devices for the tests and for the provision of advice and information. This work was supported by the Biocides and Pesticide Assessment Unit of the HSE Health Directorate.
Received March 6, 2002; in final form January 13, 2006
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL EQUIPMENT AND...ANALYTICAL METHODSSTATISTICAL METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
European Community. (1998) EC Biocides Directive 98/8/EC: Directive of the European Parliament and of the Council (EC) No 98/98 of 16 February 1998 concerning the Placing of Biocidal Products on the Market. Official Journal of the European Community(Office for Official publications of the European Communities, Luxemburg) Vol. L123:.
In Hinds WC (Ed.). Aerosol technology: properties, behavior, and measurement of airborne particles. (1999) John Wiley & sons.
Lai ACK, Byrne MA, Goddard AJH. (2002) Experimental studies of the effect of rough surfaces and air speed on aerosol deposition in a test chamber Aerosol Sci Technol. 36: pp. 973–82.
MDHS 73. (1992) Measurement of air change rates in factories and offices. (HSE Books, London, UK) ISBN 0-11-885693-6.
MDHS 82. (1997) The dust lamp—a simple tool for observing the presence of airborne particles. (HSE Books, London, UK) ISBN 0-7176-1362-3.
MDHS 94. (1999) Pesticides in air and/or on surfaces. (HSE Books, London, UK) ISBN 0-7176-1619-3; Pesticides 2000. London, UK: PSD/HSE, HMSO, 2000. ISBN 011243511.
Rowley J and Crump D. (2005) Measurements of the dispersal of aerosol sprays in a room and comparison to a simple decay model. J Environ Monitoring 7:960–3.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||