Leak Detection Using Conductivity0 pages

Application Data Sheet

Theory

ADS 43-021/rev.C
February 2009

Leak Detection Using Conductivity
BACKGROUND
Virtually all industries from food and beverage to
chemical processing use heat exchangers, condensers, or jacketed vessels. Leakage of the process
into the cooling water represents a loss of product
and can be a source of fouling or corrosion in the
cooling water system.
Conversely, leakage of the cooling water into the
process can be a source of contamination. An
example is the leakage of cooling water into steam
condensate, which is a source of contamination in
the steam-water circuit.

BASIC PRINCIPLE
The basic principle behind leak detection using
conductivity is measuring and alarming the change
in conductivity in a sample brought about by a leak
into it by a contaminant.The contaminant either
increases or decreases the conductivity of the
sample depending upon its solubility with the
process solution.
How well conductivity can detect leaks is determined
first of all by the difference in conductivity between
the sample and the contaminant. The greater the
conductivity difference, the more sensitive leak
detection will be. For any reasonable leak detection
sensitivity, the conductivity of the contaminant
should be at least 100 times greater than the sample
conductivity.
Where a specification exists for the maximum
allowable conductivity of the sample, sensitivity of
leak detection is not an issue and a conductivity
measurement is all that is required. However, even
in these instances, using two conductivity measurements can often isolate and identify the leak source.

LEAK DETECTION WITH A SINGLE
ANALYZER OR TWO ANALYZERS
(DIFFERENTIAL CONDUCTIVITY).
The second determining factor in how effective
conductivity will be in leak detection is how much the
sample conductivity must increase to reliably
indicate the presence of a leak. The smaller the

increase that is necessary, the better the sensitivity
will be.
How large a conductivity change is needed is
determined by 1) the conductivity range of the
uncontaminated sample; 2) how fast the sample
conductivity changes; and 3) the transit time, which
tells how long the sample takes to pass through the
potential leak source.
The above considerations also determine whether
there is an advantage in using differential conductivity
over using a single analyzer.

LEAK DETECTION WITH A SINGLE
ANALYZER
The conductivity of an uncontaminated sample
normally varies between a high and low extreme.
The worst case for leak detection with a single
analyzer is when the sample conductivity is at its
minimum. A leak can be present, which increases
the conductivity up to the normal sample maximum
conductivity. Since conductivity is not above a
normal, expected value, the leak is not detected.
Given this fact and the dependence of sensitivity on
the ratio of the sample conductivity to the contaminant conductivity, the overall sensitivity (Ss) with a
single analyzer can be estimated from the following:
Equation (1)
SS =

KH - KL
KC

where SS is the sensitivity in volume of contaminant
per volume of sample; KH and KL are the maximum
and minimum normal sample conductivity; and KC is
the contaminant conductivity.
Sensitivity with a single analyzer is limited by the
normal variations in sample conductivity.
Alarming for leak detection should be set at approximately 5% full scale above the maximum normal
conductivity or other reliably measurable conductivity
increase.