Common Sensing Inc. -

Notes on QA/QC procedures for TDGP sensors

Alpha Designs and Common Sensing Inc.

Guidelines drawn up by USGS, COE, USDI and NMFS which often include some confusing rationales for the methods adopted for required QA/QC procedures using TDGP sensors manufactured by our respective companies have suggested the need for some clarification from us regarding suggested methodology.

We outline briefly the advantages and drawbacks of alternative methods that have been adopted by our companies and these agencies.

Most engineering approaches to QA/QC try to meet and/or EXCEED field conditions - thereby assuring that "field conditions" are unlikely to cause failure in the particular instruments/sensors in use.

Common Sensing Inc. has always- since 1980 - applied excess pressure to the inside of the tubing by means of a hypodermic syringe. Depending on the internal volume of the membrane this can quickly add about 4-5 times atmospheric pressure to the inside of the tubing.

The Alpha Designs sensor has less internal volume than the CSI TBO-series, is more compact and as a consequence disassembly may not be desirable in the field.

Accordingly the counterdiffusion method wherein the entire probe is immersed in carbonated water constitutes an alternative approach which also produces a transient excess pressure inside the tubing as is shown in Figure 1.

Figure 1

This excess pressure can reach a full atmosphere using soda water and appropriate stirring. The figure shows the results of unstirred immersion in a carbonated beverage (we dare not reveal the brand!).

Design differences in the instruments used by different agencies would suggest that manufacturer’s recommendations should be followed, unless users of the instruments have discovered inadequacies in suggested procedures.

To our knowledge this has not occurred; what has been striven for by the manufacturers as well as agency personnel, is an acceptable method that is non-destructive, reveals the presence of existing or potential leaks in the tubing and is quickly carried out. After all, it is the manufacturers who have the most direct – and abiding - interest in insuring that the methods they recommend are satisfactory!

Both methods described here will reveal most leaks likely to be encountered; however it stands to reason that there exists a type of incipient membrane damage that will be revealed only by a more extreme overpressure.

We understand that a desire for consistency prompts the distribution of recent guidelines; however it makes no sense to require procedures involving more time or risk unnecessarily just for the sake of consistency, when the result may in fact compromise the level of QA/QC achieved. Common sense would dictate that one adopt the procedure best suited to the particular instrument design to provide the

best QA/QC. At some point everything involves a compromise between time, cost, risk of damage and risk of missing incipient failure or damage.

In summary,

1) Dimethyl silicon membranes can have small tears or weak spots caused by algae, grit or what have you. If they are small enough 760 mm Hg excess pressure achieved by counterdiffusion may not reveal them. However it is unlikely that such borderline conditions would be common under the conditions that we understand are now being used by the above agencies and provided the membranes are frequently tested (at least monthly).

2) A truly conservative test applies a more extreme pressure test on the membrane; if it passes that test it can be assumed with greater confidence to be O.K. for the next period. There is surely no disagreement with this.

3) Counterdiffusion by sodawater applies about 7.5 psi to 15 psi (760+) mm Hg to the inside of the membrane.

4) Direct increase in volume by a syringe can add four times this pressure. Small leaks are precisely what are being looked for because they can become BIG leaks. At some point everything involves a compromise between time, cost, risk of damage and risk of missing incipient failure or damage.

5) To our knowledge there have not been cases where one method has failed over the other to reveal leaks; nor do we understand that there have been opportunities to compare each method on all instruments in use. Therefore at this stage, choice of one method over the other would seem premature. And ought to be discouraged.

This would seem to be the proper objective of adequate QA/QC procedures.