In performing air sampling in the field, the first step is to assemble a sampling train consisting of a calibrated personal sampling pump connected by a piece of plastic tubing to the collection media. The appropriate collection media for the substance of interest will be specified by the laboratory methods manual and will be a sorbent tube or a filter, depending on whether you are collecting gases and vapors or particulates. The media is fitted with a small holder with a clip that allows it to be attached to the employee's collar in the breathing zone.
Once the sampling train is assembled, you are ready to go to the plant area. Alternatively, you could carry the pump and media to the field and put the sampling train together as you are placing equipment on the person to be sampled. Personally, I find it more convenient to have the sampling train assembled ahead of time. The sample is labeled and as much of the data collection sheet as possible filled out before heading to the field. For example, known data fields such as the date, sample number, pump number and flow rate can be completed in the office. It is much more convenient to perform these tasks at a desk rather than fumbling with a clipboard in the middle of a plant where there may be leaking steam lines or grease and who knows what else dripping down from an overhead crane.
Doing as many tasks as possible before arriving at the sampling location allows me to devote full attention to what I am doing and minimizes the disruption caused by my presence. By being prepared, I can swoop in quickly, brief the employee on the sampling process if I have not already done so, place the sampling equipment, complete the missing fields such as the employee's name on my data collection form, and fill in the time the pump was turned on and description of what the employee was doing during the sampling period.
Who to Sample
Some standards, such as arsenic, establish rigorous requirements stipulating that full-shift samples for at least seven continuous hours be obtained for each job classification for each shift in each work area. Other standards such as cadmium, afford more flexibility. For example, where several employees perform the same job tasks on the same shift in the same area and the length, duration and level of cadmium exposures are similar, a representative fraction of employees may be sampled. However, this sampling must include employees expected to have the highest cadmium exposure.
Briefing Employees. Once you decide who you are going to sample, you will need to explain the sampling protocol to them so they know what's going on and why you are doing it. If you hold regular safety meetings, that forum may be a good venue to demonstrate the equipment and explain the sampling process to a number of employees at the same time. If this is not possible, you could briefly explain the process to each individual before you outfit him or her with the sampling equipment. You should generally explain why sampling is being performed and provide a brief explanation of how the equipment works, that it will remain with them for the entire shift and that they should continue to go about their normal activities wearing the equipment. They should also be instructed not to fiddle with the equipment and to seek you out if they encounter any type of problem with it.
Placing the equipment. Once the employee who is to be monitored has been briefed, he can be outfitted with the sampling train. The pump is clipped to his belt. Having a few adjustable web belts purchased from a surplus store on hand will allow you to place your pumps on employees who are not wearing belts. The media holder is clipped to his shirt collar in the breathing zone, the plastic tubing should be placed behind his back and positioned so that it does not present a hazard. Loose tubing can be lightly taped to the back of an employee's shirt with strips of duct tape. The name being sampled should be recorded on your collection form as well as the sample identification number for that employee's sample. The time the pump is turned on must also be recorded.
Before turning the employee loose, you should determine when and where you will meet the employee to retrieve the equipment. Otherwise, it might go out the door with him when he leaves for the day.
Monitoring the employee's activities. It is not prudent to place equipment on an employee and then leave the area, never to be seen again. For example, you shouldn't abandon your equipment and go back to the office to make phone calls or perform paperwork. Instead, you should monitor the employee's activities so that you have a good idea what he or she was doing during the sampling period.
Sufficient information should be recorded on your data collection form to refresh your memory on what the employee was doing during the sampling period. You may also want to note any activities that might help to explain aberrations in sampling results. For example, if two employees who are nominally performing the same job have very different work practices, you should note this because it might explain a significant variation in their exposures. And if so, you may want to have the employee with the higher exposure adopt the other employee's work practices.
Other reasons for monitoring the sampling equipment are to ensure that you are readily available to address problems with the equipment should they arise and to ensure that samples are not tampered with. For example, if you are sampling for methylene chloride and the employee you are evaluating decides to place the collection media into the head space of a drum containing methylene chloride while he takes a break, the sample would be biased, indicating a higher level of exposure than the employee was actually exposed to.
Retrieving and removing the equipment. At the end of the sampling period, you will meet up with the employee. Turn off the pump, note the time it was turned off and remove the equipment.
There are three simple calculations that anyone conducting air sampling must be able to perform. They are calculating elapsed time, sample volume and an 8-hour time-weighted average.
Calculating elapsed time. While this is not complicated, you have to remember that time is not based on a decimal system. There are 60 seconds in a minute and 60 minutes in an hour, not 10 and 100. This fact must be remembered when calculating the total sampling time. For example, if a pump is turned on at 1:20 p.m. and turned off at 4:10 p.m., we calculate the elapsed time by subtracting the time we turned the pump on from the time we turned it off. We must remember that these figures represent one hour and 20 minutes and 4 hours and 10 minutes respectively, not just the numbers 120 and 410. To illustrate, 410-120=290 but 4:10 pm-1:20 pm= 2 hours and 50 minutes or 170 minutes.
Calculating sample volume. The lab must be provided with the sample volume to calculate the concentration of a contaminant. In reality, the laboratory determines the mass of the substance collected. For example, the lab may determine that there are 3 milligrams of lead on a filter used for evaluating a worker's exposure to welding fume but 3 milligrams is a mass, not a concentration. A concentration is expressed in units of a mass per volume of air. Assume that a sample volume of 700 liters was required to collect this 3 milligrams of lead, and take it on faith that 1,000 liters of air is the same as one cubic meter of air. Thus, 700 liters is the same as 0.7 cubic meters.
To determine the concentration of lead in the air, the laboratory will divide the quantity of lead that it measured in the sample by the sample volume. So in this case, 3 mg divided by 0.7 cubic meters yields a concentration of 4.28 milligrams per cubic meter. The laboratory will usually make this calculation for you and report the results in milligrams per cubic meter for particulates or parts per million for gases and vapor, but to make that calculation, the laboratory must know the volume of air sampled and it is our responsibility to provide that. If we report an inaccurate air volume, the results obtained from the laboratory will also be inaccurate, so you can see the importance of proper pump calibration and accurately noting the time the sampling pump was turned on and off.
The volume of air sampled is obtained by multiplying the flow rate by the duration of the sampling. For example, if the pump flow rate is 2 liters per minute and we sample for 2 hours (i.e. 120 minutes), the sample volume is determined by multiplying the flow rate by the sampling time. In this case, 2 lpm x 120 minutes is 240 liters. If instead we sampled for 3 hours (i.e. 180 minutes), the sample volume would be 360 liters. Note that you must be careful to maintain consistent time units so the flow rate in liter per minute is multiplied by the sampling time in minutes, not the sampling time in hours. So a 2-hour sample is 120 minutes, 3 hours 360 minutes and so forth.
Calculating 8-hour time-weighted averages. Anyone doing air monitoring should be able to calculate an 8-hour time-weighted average and there may be some occasions where you will have to do this. If only one sample is collected and it is collected over an 8-hour period, the laboratory results for that sample constitutes an 8-hour time-weighted average but suppose 6 hours into sampling for welding fumes, you are checking your equipment and the sampling line pops off the pump inlet. You remove the filter and replace it with a second filter noting the difference between the two samples by giving each a different sample number and making appropriate notes on your data collection form. Your stopping time for the first sample is essentially the starting time for the second sample. You continue sampling for the next 2 hours without further incident.
When you receive your results, the lab reports a concentration of 10 milligrams per cubic meter for the first sample and 20 milligrams per cubic meter in the second sample. Since the samples were collected over differing lengths of time 6 hours for the first and 2 hours for the second you cannot perform a simple arithmetic average. You can't take 10 milligrams per cubic meter and 20 milligrams and divide by 2 to obtain an average of 15 milligrams. Instead, you must weight each sample for the period of time over which it was collected. Although this may sound complicated, it isn't. You simply take the first concentration and multiply it by its sampling time and add it to the second concentration time's first sampling time and divide all of this by the total sampling time. If there are more samples taken, you continue in a similar fashion expressed symbolically as an equation. The formula looks like this:
[(C1T1)+(C2T2)+(C3T3)]/T total = 8-hour time-weighted average
Where C1 is the concentration of first sample: in our example, 10g/m3
T1 is sampling time for the first sample: in our example, 6 hours
C2 is the concentration of the second sample: in our example 20 mg/m3
T2 is the sampling time for the second sample 2 hours
So our example yields:
[(C1T1)+(C2T2)]/total time = [(10x6)+(20x2)]/8 hours = [(60)+(40)]/8 = (100)/8 = 12.5 mg/m3
Note that 8-hour TWA of 12.5 mg/m3 is a much different than the simple arithmetic average of 15 mg/m3
While many sampling situations demand the experience and skill of a certified industrial hygienist, sampling that is repetitive and routine such as that required by OSHA's substance-specific health standards can be performed by someone with a little mechanical aptitude and dexterity. Although OSHA's substance-specific standards have many requirements in common, minor differences that exist among the standards demand that anyone performing sampling to meet the requirements of the standard be intimately familiar wit the specific regulatory requirements for the substance they are sampling. Safety practitioners who for one reason or another feel intimidated about doing air sampling can develop confidence and skill by attending a formal short course on sampling or arrange for private tutoring from a local industrial hygienist.
Contributing Editor John F. Rekus, PE, CIH, CSP, is an independent consultant. He has more than 20 years of regulatory experience and may be reached at (410) 583-7954.