Which Reference Electrode Goes in the Truck
Before a survey, a technician checks two copper/copper-sulfate reference electrodes against a third electrode kept as a shop reference. Electrode A reads 5 mV and electrode B reads 19 mV against the shop cell. Electrode B's viewing window shows a pale, nearly clear solution with only a few crystals left; electrode A's solution is dark blue with crystals still visible above the liquid. Both electrodes spent the night in the truck; the morning air temperature is 34 degrees C and B was topped up with tap water on site the week before. Only one electrode can be carried on the day's 4 km close-interval survey, and the acceptance limit in the company procedure is 5 mV.
1. Which electrode should be used, and why does a 19 mV offset matter more than it looks?
Read the worked answer
Use electrode A. It sits exactly at the procedure's 5 mV acceptance limit; B is nearly four times the limit.
Why the size of the offset matters: a reference-electrode offset is additive to every reading taken with it. On a station measuring -0.845 V CSE, a 19 mV error can push the reported value across the -0.850 V criterion either way. That converts a measurement problem into a compliance problem, and the error is systematic, not random: it will not average out over 800 readings.
What B's appearance tells you: pale, nearly clear solution with few crystals means the cell is not saturated with copper sulfate. An unsaturated cell has a drifting, concentration-dependent junction potential, which is consistent with the 19 mV offset. Topping up with tap water introduces chloride and other ions that further shift the potential and can foul the porous plug.
Caution on direction: before using any correction, establish the sign of the offset (which cell is positive) on a known pair. A correction applied with the wrong sign doubles the error instead of removing it. Since A is at the limit rather than comfortably inside it, note the check result in the day's records.
2. What is wrong with electrode B and how would you restore it and prove it is fit to use?
Read the worked answer
Faults to look for, in order:
- Not saturated - too few crystals; the cell needs excess undissolved crystals at all times.
- Wrong water - tap water added on site dilutes the copper sulfate and contaminates the cell.
- Plug condition - a dried, clogged or cracked porous plug raises the cell's resistance and produces unstable readings.
- Trapped air - a bubble against the plug breaks the ionic path.
- Temperature - both electrodes were in a hot truck overnight; a cell should be at ground temperature before comparison.
Restoration: empty and rinse, refill with distilled or deionized water and reagent-grade copper sulfate with excess crystals, tap out bubbles against the plug, soak upright for at least a day, and clean or replace the plug.
Proof: re-compare against the shop reference cell and against electrode A, after both have equilibrated to the same temperature. Only if it comes back inside 5 mV is it fit for survey work.
Do not put a restored cell straight onto a survey. The restoration itself has to be a recorded, verified event.
3. The technician says checking A against B against the shop cell is enough. What is the weakness in that logic, and how should the check be scheduled during the day?
Read the worked answer
The weakness: two electrodes compared only to each other cannot reveal a common-mode drift. If both have drifted in the same direction, the difference between them stays small and looks like a pass. The shop cell is the only independent witness - which is exactly why the third cell must be genuinely independent and itself checked periodically, not just kept in a drawer.
Practical consequences:
- A pass on the A-vs-B difference is not evidence that either is correct; it is evidence they agree.
- Any field correction is only defensible if traceable to the independent cell.
Scheduling the checks:
- Before the first reading and after the last reading of the day.
- After any large temperature change, long transport, or a drop.
- Any time a reading looks out of trend with its neighbours.
If a check mid-survey shows drift, the readings taken since the last good check are suspect. The correct response is to flag that block of data for re-survey, not to apply a guessed average correction after the fact.