Varied results between different brands of Sampling kits
I have a consultant who recently started using QED's MicroPurge kit, they had previously used Waterra foot valves. They have seen a marked increase in the levels of heavy metals after changing to the MicroPurge kit. Can you think of any reason why this could be? Also, if this makes any difference, they were field filtering with stericups while previously they were sending the samples to the lab to be filtered before processing. Any help would be appreciated.
Actually, this isn't as strange as it first seems, and it is something we have seen before with customers who have switched from sampling methods such as bailers or inertial lift pumps (i.e., Waterra pumps) to low-flow sampling with bladder pumps. The client has also switched from sending their samples to the lab for filtration to filtering in the field with Millipore Stericup filters, which are a vacuum filtration bottle and cup assembly with an integral membrane filter pad holder. This filtration change adds another wrinkle – the two issues will be covered separately.
In cases where groundwater is purged and sampled with an inertial lift pump, we expect to see elevated sample turbidity. This elevated turbidity can result in elevated concentrations for many metals in unfiltered or "total" samples ("total" referring to the analysis of both the dissolved metal ions and any metals associated with suspended solids in the sample). By switching to low-flow sampling, we expect to see reduced sample turbidity, which should then result in reduced total metals concentrations. However, dissolved (filtered) metals concentrations can often go UP at the same time that total concentrations are reduced. This is because the bladder pump is better at maintaining metal ions in solution since there is no air contact with the sample and minimal sample agitation that could result in sample aeration (oxidation) and precipitation of metals out of solution. In effect, a low flow sample provides a better estimate of both the true level of suspended solids in the groundwater as well as a more accurate value for the true level of dissolved metals than bailing or inertial lift pumping. In cases where regulatory agencies are concerned about the higher reported values, we have been successful in providing this explanation to them and obtaining a "grace period" to reestablish baseline groundwater conditions where statistical data comparisons are performed.
Changing filtration methods can have an even greater impact on metals concentrations. First some background on ground-water sample filtration: Turbid samples are commonly filtered to remove the suspended solids but, unfortunately, the common filter pore size used is 0.45-micron. This pore size doesn't really remove all solids that are suspended; colloidal particles down to 0.1-micron are often seen in groundwater samples. Also, filtration is not an absolute process, that is, the stated pore size of the filter doesn't dictate the size of what passes through the filter, but rather the particle size that will not pass through. In the case of a 0.45-micron filter, it is common to see particles removed down to 0.3-0.2 microns in size, and this will vary with particle shape, attraction to the filter membrane material, the buildup of a "filter cake" on the membrane surface that traps smaller and smaller particles as it builds up. So, where samples from bailers or inertial lift pumps are highly turbid, the effect of the greater filter cake buildup can end up removing a higher percentage of smaller particles, whereas filtration of the bladder pump sample with much lower turbidity can actually allow more small particles below pore size to pass through the filter, resulting in higher "dissolved" metals concentrations.
There is a second filtration issue with changing over from lab filtration to field filtration. Where samples are collected for metals analyses and then filtered in the field, the common practice is to add a nitric acid preservative to the sample to reduce the pH to 2.0 or less, which maintains the dissolved metals concentrations during sample shipment and prior to analysis. Where samples are to be filtered in the lab, the common practice is to collect the sample in a plastic bottle and then send it to the lab without the addition of any nitric acid preservative to avoid a high bias in dissolved metals concentrations where those metals are present on or in the suspended solids in the sample. Unfortunately, this lack of sample preservation with acid can then result in the metals that are in solution precipitating out due to exposure of the groundwater to well head conditions where it can rapidly gain dissolved oxygen and slowly lose dissolved CO2, both of which can reduce the dissolved metals concentrations. This can sometimes be observed in cases of samples with very high dissolved iron, where the groundwater in the sample bottle initially looks clear but becomes reddish brown over a period of minutes or hours due to precipitation. Since iron can "co-precipitate" some other metals out of solution, there can be a 'snowball effect' on other dissolved metals concentrations. The preferred method is to sample in the field and then preserve the samples immediately, as they appear to be doing now.
Finally, there is an issue with the type of filter mechanism and the filter media materials. If the Stericup filter membranes are made of a different material than whatever the lab was using (e.g., PVDF or PES membrane Stericup versus glass fiber or acrylic membranes at the lab), this could affect results. Also, the Stericup is a vacuum filtration vessel, which can also cause some change in dissolved metals concentrations due to exposure of the sample to air as well as the vacuum (negative pressure) that can more rapidly remove dissolved CO2.
QED’s recommendation would be to stick with the Sample Pro pump and use proper low-flow procedures to control turbidity, and then filter samples for dissolved metals in the field using an in-line filter capsule such as QED's QuickFilter capsules. QuickFilters would be competitive in price with the Millipore Stericup.
Actually, this isn't as strange as it first seems, and it is something we have seen before with customers who have switched from sampling methods such as bailers or inertial lift pumps (i.e., Waterra pumps) to low-flow sampling with bladder pumps. The client has also switched from sending their samples to the lab for filtration to filtering in the field with Millipore Stericup filters, which are a vacuum filtration bottle and cup assembly with an integral membrane filter pad holder. This filtration change adds another wrinkle – the two issues will be covered separately.
In cases where groundwater is purged and sampled with an inertial lift pump, we expect to see elevated sample turbidity. This elevated turbidity can result in elevated concentrations for many metals in unfiltered or "total" samples ("total" referring to the analysis of both the dissolved metal ions and any metals associated with suspended solids in the sample). By switching to low-flow sampling, we expect to see reduced sample turbidity, which should then result in reduced total metals concentrations. However, dissolved (filtered) metals concentrations can often go UP at the same time that total concentrations are reduced. This is because the bladder pump is better at maintaining metal ions in solution since there is no air contact with the sample and minimal sample agitation that could result in sample aeration (oxidation) and precipitation of metals out of solution. In effect, a low flow sample provides a better estimate of both the true level of suspended solids in the groundwater as well as a more accurate value for the true level of dissolved metals than bailing or inertial lift pumping. In cases where regulatory agencies are concerned about the higher reported values, we have been successful in providing this explanation to them and obtaining a "grace period" to reestablish baseline groundwater conditions where statistical data comparisons are performed.
Changing filtration methods can have an even greater impact on metals concentrations. First some background on ground-water sample filtration: Turbid samples are commonly filtered to remove the suspended solids but, unfortunately, the common filter pore size used is 0.45-micron. This pore size doesn't really remove all solids that are suspended; colloidal particles down to 0.1-micron are often seen in groundwater samples. Also, filtration is not an absolute process, that is, the stated pore size of the filter doesn't dictate the size of what passes through the filter, but rather the particle size that will not pass through. In the case of a 0.45-micron filter, it is common to see particles removed down to 0.3-0.2 microns in size, and this will vary with particle shape, attraction to the filter membrane material, the buildup of a "filter cake" on the membrane surface that traps smaller and smaller particles as it builds up. So, where samples from bailers or inertial lift pumps are highly turbid, the effect of the greater filter cake buildup can end up removing a higher percentage of smaller particles, whereas filtration of the bladder pump sample with much lower turbidity can actually allow more small particles below pore size to pass through the filter, resulting in higher "dissolved" metals concentrations.
There is a second filtration issue with changing over from lab filtration to field filtration. Where samples are collected for metals analyses and then filtered in the field, the common practice is to add a nitric acid preservative to the sample to reduce the pH to 2.0 or less, which maintains the dissolved metals concentrations during sample shipment and prior to analysis. Where samples are to be filtered in the lab, the common practice is to collect the sample in a plastic bottle and then send it to the lab without the addition of any nitric acid preservative to avoid a high bias in dissolved metals concentrations where those metals are present on or in the suspended solids in the sample. Unfortunately, this lack of sample preservation with acid can then result in the metals that are in solution precipitating out due to exposure of the groundwater to well head conditions where it can rapidly gain dissolved oxygen and slowly lose dissolved CO2, both of which can reduce the dissolved metals concentrations. This can sometimes be observed in cases of samples with very high dissolved iron, where the groundwater in the sample bottle initially looks clear but becomes reddish brown over a period of minutes or hours due to precipitation. Since iron can "co-precipitate" some other metals out of solution, there can be a 'snowball effect' on other dissolved metals concentrations. The preferred method is to sample in the field and then preserve the samples immediately, as they appear to be doing now.
Finally, there is an issue with the type of filter mechanism and the filter media materials. If the Stericup filter membranes are made of a different material than whatever the lab was using (e.g., PVDF or PES membrane Stericup versus glass fiber or acrylic membranes at the lab), this could affect results. Also, the Stericup is a vacuum filtration vessel, which can also cause some change in dissolved metals concentrations due to exposure of the sample to air as well as the vacuum (negative pressure) that can more rapidly remove dissolved CO2.
QED’s recommendation would be to stick with the Sample Pro pump and use proper low-flow procedures to control turbidity, and then filter samples for dissolved metals in the field using an in-line filter capsule such as QED's QuickFilter capsules. QuickFilters would be competitive in price with the Millipore Stericup.