METHOD AND APPARATUS FOR MEASURING A PROPERTY OF AIR
Field of the Invention
The present invention relates to a method of calculating a measure of a property of air and apparatus for measuring a property of air. The invention is particularly suited to measuring the concentration of particles of different sizes in air.
Background to the Invention
Current techniques for measuring concentrations of particles in air for air quality assessment involve the use of particle measurement systems. Such particle measurement systems have a selective size inlet (SSI) which selectively allows particles smaller than the cutoff size to pass the inlet so that properties of that particle size fraction can be measured. Typically, a SSI allows particles of less than one of the sizes lOμm, 2.5μm or lμm to pass. The particle measurement system has an air pump to cause air to flow through the SSI to a icrobalance (such as the Tapered Element Oscillating Microbalance available from Rupprecht and Patashnick Inc.) which measures the mass of particle material collected from the air flow in real time. The measured mass is then used to produce a measure of the concentration of particles in the air. These measurements are sent to a data logger which stores the information for further processing.
Measurements of more than a single particle size are made by co-location of separate particle measurement systems for each particle size being measured. The capital cost of the overall particle measurement system is dominated by the cost of the microbalance, pump, supporting electronics and data logger, with the selective size inlet being a minor cost component.
The inventor has realised that because levels of
particles in the atmosphere such as air pollutants, change slowly, if data is gathered over a plurality of sample periods spaced throughout the measurement period, it is possible to obtain valid data over the measurement period without measuring constantly.
Summary of the Invention
Accordingly, the invention provides apparatus for measuring a property of air for two different particle sizes over a measurement period, said apparatus including: a first size selective inlet; a second size selective inlet with a cut-off size different to said first size selective inlet; a shared measurement device for measuring said property of air; and air flow directing means for directing air through one of said first and said second size selective inlets at a time to said shared measurement device, the air flow directing means being configured to repeatedly switch air flow during said measurement period so that the air flow is directed through each of said first and second size selective inlets for a plurality of sample periods to thereby obtain a plurality of sample measures for each of said first and second size selective inlets which can be used to calculate a measure of the property for each of said particle sizes, wherein the air flow directing means is configured to switch said air flow such that any change in the property of air between consecutive sample periods for each particle size will not significantly affect calculated measures of the property of air for the measurement period.
Preferably, said shared measurement device is used to obtain sample measures of the concentration of particles in air.
Preferably, said air flow directing means includes a controllable valve located between both the
size selective inlets and said shared measurement device for controlling which of the size selective inlets air is directed through.
Preferably, said airflow directing means includes an air pump for directing air through said size selective inlets .
Preferably, said particle sizes are chosen from the group of: particles less than lOμm in size; particles less than 2.5μm in size; and particles less than l.Oμm in size, where size is defined as aerodynamic diameter.
Preferably, said sample period is in the range of 2-15 minutes.
Preferably, said sample period is about five _. minutes.
Thus, when two particle sizes are being measured, their respective sample periods will typically be spaced apart by five minutes .
Alternatively, said sample period is about ten minutes .
Preferably, said measurement period is between 1 and 48 hours.
Preferably, said measurement period is about 24 hours .
Alternatively, said measurement period is about one hour .
In some embodiments, consecutive sample periods
may be spaced by switching periods.
The invention also provides a method of calculating a measure of a property of air for a plurality of different particle sizes over a measurement period, said method involving: measuring said property of air for each different size during a plurality of spaced apart sample periods by measuring the property of air for each of said different sizes in turn so that the sample periods of respective ones of said particles sizes are spaced apart by sample periods of the other particle size or sizes; and calculating said measure of the property of air for each of said particle sizes from the measurements obtained during said sample periods, wherein the duration and spacing of consecutive sample periods for each different particle size are chosen such that any change in the property of air between consecutive sample periods does not significantly affect the calculated measure of the property of air over the measurement period.
Brief Description of the Drawings
A preferred embodiment will now be described with reference to the accompanying drawings in which:
Figure 1 is a block diagram showing the apparatus of the present invention;
Figure 2 is a graph showing calculated and measured data.
Description of the Preferred Embodiment
The present invention arises from the realisation that levels of particles in the atmosphere, and in particular air pollutants, change slowly through a diurnal pattern. This relatively slow variation in the diurnal pollutant patterns is the reason why regulatory authorities typically publish data at hourly-average or longer averaging times. Indeed, statutory authorities
world-wide mandate PM10 (particles of less than lOμm) and PM2.5 (particles less than 2.5μm standards only as 24 -hour averages. Thus, accuracy at these long averaging times is ultimately what is required of measuring systems.
The invention described in the preferred embodiment was made by the realisation that against these long averaging times, and the slowly-varying nature of PM levels valid PM data at the hourly or mandated daily- average level should be obtainable using only sample periods across the measurement period. For example, in the preferred embodiment every second or third five-minute period in each hourly or daily measurement period is used as a sample period during which the concentrations of particles in air is measured for particles of a particular size.
This allows, in the apparatus of the preferred embodiment, a single microbalance to measure air which has been directed through two or more size selective inlets while still allowing the estimation of valid hourly and daily-average data from the data collected during the sample periods for each of the particle sizes.
Figure 1 is a block diagram of the apparatus of a preferred embodiment of the present invention.
A pump (not illustrated) is operable to cause air to flow through three size selective inlets which are provided by a first size selective inlet (SSI) 1 which only allows particles of size PM10 or less to pass (i.e lOμm or less) . A second SSI 2 only allows particles of size PM2.5 or less to pass (i.e.2.5μm or less), and a third SSI 3 only allows particles of size PM1 or less to pass (i.e. lμm or less). Herein the term "size" refers to the aerodynamic diameter of the particles.
The outlets of each of the SSIs are in fluid
communication with a low particle-loss valve 4 which is controlled to direct each of the air flows from the respective SSIs in turn to a Tapered Element Oscillating Microbalance (TEOM) 5 (available from Rupprecht and Patashnick Inc.). Thus, in combination the pump and the valve provide means for directing air flow through each of the size selective inlets to a shared microbalance.
In the preferred embodiment, the valve 4 is programmed so as to direct the air flows of the respective SSI's in turn for a sample period of 5 minutes. The microbalance weighs the particles in the air flow during this sample period to produce a real-time measure of the concentration of particles in the air. The concentration measure for the sample periods are then communicated to the data logger 6 which stores the concentrations for future or real-time processing. The logged data can then be processed to provide measures of the concentrations of particles of each of the three sizes during the measurement period. The measurement period is typically one hour or 24 hours depending on the application as these are the periods used by statutory authorities. It will be apparent to persons skilled in the art that other measurement periods can be used.
Thus, the capital costs of such a particle measurement system is greatly reduced relative to the current situation where it is necessary to co-locate multiple particle measurement systems to measure concentrations of particles of different sizes because it is possible to sample PM10, PM2.5 and PMl data using a single microbalance which is the major cost of such measurement systems .
Further, it is possible to improve data quality by use of a single particle measurement system rather than multiple co-located systems as this eliminates calibration differences.
Performance potential for this system was tested via two numerical simulations based on eight days of raw PM10 data obtained using a tapered element oscillating microbalance (TEOM) and sample periods of 5 minutes . (TEOM is a registered trade mark of Rupprecht and Patashnick, Inc) . The data was recorded at an urban site in the city of Melbourne, Australia. In the first simulation every alternate data point was deemed to be from a PM2.5 SSI, with its value assumed to be a factor of 0.7 times the original PM10 value. This generated two data series, one PM10 series consisting of half the original 5-minute values (i.e. every second data point), and a fictitious PM2.5 series consisting of the other half of the original PM10 series, scaled by a factor of 0.7. Thus each data series consisted of one 5-minute data point taken (alternatively) every ten minutes.
Complete data serdes were reconstructed from these two series by using cubic spline interpolation to provide an interpolated PM10 or PM2.5 value for each of the, alternate, missing 5-minute data points. A comparison between these reconstructed data series and the original, complete PM10 data series, and fictitious PM2.5 data series (the measured PM10 series scaled by 0.7) provides a test of the concept.
The second simulation was analogous, but was based on time-sharing the measurements across three SS s, PM10, PM2.5 and PM1, each measurement being allocated to very third 5-minute data point. As in the previous simulation PM2.5 was assumed to equal PM10 scaled by 0.7. PM1 was assumed to equal PM10 scaled by 0.4.
Performance of the concept was tested for each simulation by computing and comparing hourly and daily averages from the original complete data series and the reconstructed data series derived from the time-shared
series. Figure 2 shows the comparison for hourly data points . Table 1 contains regression results from the hourly data comparison. Table 2 shows the errors in daily average results that result from the simulations.
Table 1. Results of regression of hourly-average data from reconstructed data series versus complete original data series (192 data pairs) .
Table 2. Errors in 24-hour average PM10 (μgm~3) calculated from reconstructed PM10 data series yielded by numerical simulations with the EROM time-shared between two or three separate inlets .
The result of the simulations suggest a negligible difference in hourly and daily-averaged data derived from the time-shared series, in comparison with the original complete series .
It will apparent from the foregoing, that the apparatus of the preferred embodiment can be modified so as to incorporate two SSIs or alternatively that only two SSIs need to be used at any one time. Further, additional SSIs may be provided if it is desired to measure particles of other sizes. For example, it will be apparent that there may be a switching period between consecutive sample periods during which the valve 4 is switched from one SSI to another and this can be accounted for in the calculation of the measure of particle concentrations.
The sample period of five minutes is chosen for convenience as this is a standard measurement period of the TEOM. It will be apparent to persons skilled in the art that other sample periods can be used.
It will be apparent to persons skilled in the art, that while the present technique is particularly suited to measurement of particle concentrations, it may also be used to measure other properties of air which vary over time. For example, measurement of the optical property Bsp (aerosol scattering coefficient) for particles in selected size ranges rather than particle mass concentration could be achieved by replacing the microbalance (TEOM) in the embodiment with an appropriate optical measurement system (a nephelometer) .
Various other modifications will be apparent to persons skilled in the art and should be understood as falling within the scope of the invention described herein