General discussion on sorption
Table 5. Relationship between the pumping
time (t) and the relative concentration (Ct)
Our low-flow-rate studies indicate that al-
of TCE in water pumped through 100 ft of
though the more sorptive plasticized polypropy-
tubing at different flow rates.
lene tubings do reach equilibrium with the con-
taminated well water, the change in the normal-
Contact
Flow
time
(min)*
r†
(mL/min)
Relationship
enough to bias results. It appears that the less
sorptive LDPE could be used to sample TCE in all
LDPE
but the deepest wells (> 100 ft) if the tubing were
100
10
Ct = 0.785 + 0.0932 log(t)
0.954
equilibrated (pumped) for 24 hr. The question
1,000
1
Ct= 0.986 + 0.0082 log(t)
0.706
that then remains is what would happen with the
PP1
other analytes that are more readily sorbed by
100
10
Ct = 0.492 + 0.102 log(t)
0.909
1,000
1
Ct = 0.941 + 0.0205 log(t)
0.811
this tubing, such as PCE and PDCB (Parker and
Ranney 1996). Based upon our previous findings
* Time for water to pass through tubing.
(Parker and Ranney 1996), we expect that losses
† Correlation coefficient.
of PDCB and PCE would be substantially greater
Table 6. Predicted normalized
concentrations of TCE following
However, currently this is unproven.
pumping at 1 L/min for various
Clearly, the fluoropolymers tested in this
pumping times.
study [PVDF, P(VDF-HFP)] did not sorb much
TCE. We were surprised to find there was no sig-
Sample
nificant difference between the concentrations of
Tubing
contact
Pumping Predicted
length
time
time
normal
TCE in samples exposed to the PVDF and P(VDF-
(ft)
(min)*
(hr)
conc. TCE
HFP) tubings. In our previous study (Parker and
Ranney 1996), losses were always much greater
100
1
4
0.98
500
5
4
0.90
for test solutions exposed to the flexible fluo-
1,000
10
4
0.84
ropolymers than those exposed to the rigid fluo-
100
1
2
0.98
ropolymers. We suspect that we would have de-
500
5
2
0.88
tected some differences between these two mate-
1,000
10
2
0.81
rials if a longer length of tubing or a more sorp-
* Time for water to pass through tub-
tive analyte had been tested. However, this also
ing.
remains unproven.
ft of LDPE at 1 L/min (Table 5) and water
If the samples are to be pumped at a rapid rate
pumped through 10 ft of LDPE tubing at 100
(1 L/min), it appears that LDPE tubing can be
mL/min (Table 4). In both cases, the contact time
used to sample most wells for TCE, although 24
was 1 min. Thus, it appears that contact time can
hr should be allowed for equilibration in wells
be used to predict losses.
500 ft and greater. However, additional tests
We used the relationships derived in Table 4
should be conducted with longer lengths of tub-
for 5- and 10-min contact times to predict what
ing and more sorptive analytes to prove this.
the losses might be for TCE-contaminated water
pumped through 500 ft and 1,000 ft of LDPE at 1
Leaching findings
L/min, after allowing the tubing to equilibrate
We did not detect any spurious peaks, using
for either 2 or 4 hr. We see that if only 2 hr were
RP-HPLC analysis and a UV detector, in any of
allowed for equilibration, the estimated losses
the samples that had been pumped through any
would be only slightly greater than 10% for 500 ft
of the tubings used in these experiments, even
(Table 6). Only for the 1000-ft lengths would a
the longest lengths. However, we should men-
longer equilibration time (4 hr) be required if the
tion that because the concentration of our analyte
flow rate was 1 L/min. Thus, we would predict
of interest, TCE, was in the low mg/L range,
that longer lengths (500 ft and greater) of LDPE
these analyses were not highly sensitive. Also,
could be used to sample TCE at a flow rate of 1
UV detectors are not universal detectors and
L/min if 24 hr is given for equilibration. How-
many organic constituents that might be leached
ever, these predictions need to be verified by ad-
would not be found using this type of detector.
ditional testing, and other analytes need to be
UV detectors are most sensitive for conjugated
considered.
bonds and sulfur-, nitrogen-, bromine-, and iodine-
12
Previous Page
