pH, ORP and analytical probes that drift

Why is my pH probe slope low after calibration, and what slope percentage and zero offset are acceptable?

Acceptable slope is 95 to 105 percent of theoretical, with a zero offset inside ±30 mV. A low slope is stale buffer, a coated bulb, or a dried-out bulb far more often than it is a dead electrode.

Theoretical slope at 25 °C is 59.16 mV per pH unit. A perfect electrode moves 177.5 mV between pH 7 and pH 4 buffer. Slope also moves with temperature — 54.20 mV/pH at 0 °C, 58.17 at 20 °C, 64.12 at 50 °C — so a meter doing the math at the wrong temperature reports a false slope from a good probe.

• 95 to 105% (56.2 to 62.1 mV/pH): pass.
• 85 to 95% (50.3 to 56.2 mV/pH): clean the bulb, rehydrate in 3 M to 4 M KCl, open fresh buffer, recalibrate.
• Below 85% (under 50.3 mV/pH) after cleaning and fresh buffer: replace the electrode. Plenty of sources draw the replacement line higher, at 90% (53.2 mV/pH), so a probe in the 85 to 90 band is arguably already finished.
• Zero offset beyond ±30 mV in pH 7 buffer: the reference is plugged or poisoned, not the glass. Some procedures tighten this to ±20 mV.

Those bands are the common default. Where a transmitter, an SOP, or a manufacturer states its own acceptance limits, those govern instead.

Work the causes in that order: buffer past its open date, coating on the glass, a bulb stored dry, a plugged reference junction, then the electrode itself. Skip the order and you buy a new electrode, drop it into the same conditions, and watch it read low again in the same number of weeks.

What causes a pH reading to drift over time?

Steady drift in one direction is the reference electrode, not the glass. The glass sets slope. The reference sets zero, and a poisoned or plugged reference moves the whole curve.

Sulfide, cyanide, bromide, iodide and heavy metals attack the Ag/AgCl reference element. Process solids plug the junction. In a refillable cell the fill solution depletes, or process ingress dilutes it. A few mV per week is normal aging. Tens of mV in a single shift is contamination, a plugged junction, or lost fill solution.

The other common cause is temperature compensation, and it hides well, because the error grows with distance from pH 7. A 10 °C compensation error costs about 0.10 pH at pH 10 and about 0.17 pH at pH 12, and nothing at all at pH 7. Pull the probe and read fresh buffer. If it calibrates clean and then drifts again in service, the process is attacking the reference and the transmitter is innocent.

Why does my pH probe read correctly in buffer but wrong in the process?

The electrode is fine. The installation is not. Buffer is ionically rich and electrically quiet. The process is neither.

Four things cause this, and buffer reproduces none of them:

• Liquid junction potential. Water below about 100 µS/cm can generate tens of mV across the reference junction. Use a flowing-junction or low-ionic-strength electrode.
• Ground loops. A stray current path through the process shows as an offset that changes with pump or agitator state. A solution ground (liquid earth) electrode and single-point instrument grounding fix it.
• Sodium error. Alkali metal ions displace hydrogen ions in the gel layer, so the electrode reads low in high-sodium caustic solutions. Where it starts depends on the glass: older sodium-based formulations showed it from around pH 9 to 10, while modern lithium glass often shows nothing until pH 12 to 13. The error grows with pH, with sodium concentration, and with temperature, and in strong caustic it runs from a few tenths of a pH unit to more than a full unit. High-pH glass is a different part number.
• Temperature. Buffer at 25 °C, process at 80 °C, and an unmatched or failed Pt100/Pt1000 element.

One test separates them. Pull a grab sample into a plastic beaker and read it with the same probe, out of the pipe. Beaker right, in-line wrong, same water: the fault is grounding or junction potential in the installation, not the sensor.

Can you store a pH probe in distilled or DI water?

No. DI and distilled water leach KCl out of the reference and ions out of the hydrated glass layer. Days of that will ruin an electrode.

Store in the manufacturer's storage solution, normally 3 M to 4 M KCl. pH 4 buffer will do if that is what is on the shelf, and a 50:50 mix of KCl storage solution and pH 4 buffer is the common shop answer, close to what most manufacturers ship in the wetting cap. Never dry. Never DI, distilled, or RO permeate. The damage shows up as low slope, slow response, and a zero offset that will not come back inside ±30 mV.

A bulb that has only dried out will often come back after a few hours to overnight in KCl storage solution or pH 4 buffer. A probe that sat in DI for weeks with a depleted reference usually will not. Tap water is a poor substitute. It still beats DI, because it at least has ions in it.

How do I clean a coated or fouled pH electrode without wrecking it?

Match the cleaner to the coating. Soak, do not scrub. Nothing abrasive goes near the glass bulb.

• Oil and grease: warm water with mild detergent, or a brief rinse in isopropyl alcohol. Do not soak in solvent — organics attack the junction and some body materials.
• Scale, carbonate, hydroxide: 0.1 M HCl for 10 to 15 minutes.
• Protein: 1% pepsin in 0.1 M HCl, soaked 1 to 2 hours.
• Sulfide and silver sulfide deposits: thiourea in 0.1 M HCl, typically 30 to 60 minutes, until the dark film clears.
• Biofilm: household bleach diluted about 1:10, briefly, then rinse well.

Check the electrode manufacturer's cleaning instructions first. Acid and solvent cleaners attack some body and junction materials, and their list governs over a general one.

Never use hydrofluoric acid — it dissolves the glass bulb outright. No wire brush, no scouring pad, no abrasive of any kind. Do not wipe the bulb dry with a rag either — that scratches the glass and leaves a static charge on it that takes minutes to bleed off.

Any aggressive chemical clean strips the hydrated gel layer, and the electrode reads badly until it rebuilds. So finish the job. Rinse with DI, rehydrate in KCl storage solution or pH 4 buffer for at least 20 to 30 minutes, then recalibrate. A cleaned electrode that has not been recalibrated is not back in service.

How often should you calibrate a pH probe on a process line?

There is no universal interval. Start weekly, record the as-found error every time, and let the measured drift set the schedule.

The as-found reading is the entire point of the exercise. Put the probe in fresh buffer, write down what it reads, and only then adjust. That number is your drift rate. The interval is right when the as-found error stays inside the tolerance the process actually needs. Clean utility water often holds 4 to 8 weeks. Hot, dirty, or sulfide-bearing service can need weekly or more often.

Use two points that bracket the control point: pH 7 and pH 4 for acid service, pH 7 and pH 10 for caustic. Calibrate at process temperature where that is practical. Under an NPDES permit, a drinking water rule, or cGMP, the permit or the SOP sets the interval and a rule of thumb does not override it.

Do pH buffer solutions expire once you open the bottle?

Yes. A common manufacturer rule is 30 days for pH 10 and 90 days for pH 4 and pH 7 after opening, with unopened bottles rated 1 to 2 years. The printed date on the label governs over any rule of thumb.

pH 10 goes first. It absorbs CO2 out of the air — air runs about 0.04% CO2 — and that forms carbonic acid and pulls the buffer value down. Stale pH 10 makes a perfectly good electrode calibrate to a low slope. This is why the bad probe so often turns out to be a bad bottle.

Pour buffer into a clean cup, use it once, throw it out. Never dip the probe into the stock bottle and never pour used buffer back. Write the open date on the bottle in marker, and read the printed expiration date while you are there. Published windows vary widely between suppliers. 2 years unopened for pH 4 and pH 7 with 90 days after opening is one common statement; some suppliers rate unopened stock at 18 months, or guarantee only 12 months from shipment. pH 10 is always the shortest of the three. The label wins.

What is the difference between ORP and pH measurement?

pH measures hydrogen ion activity on a calibrated scale, in pH units. ORP measures the net oxidizing or reducing tendency of the whole solution, in raw millivolts, and it has no true calibration.

The hardware looks nearly identical. Same body, same Ag/AgCl reference. The measuring element is pH-sensitive glass on one and a platinum or gold band on the other. pH converts millivolts to pH units through a slope, 59.16 mV/pH at 25 °C. ORP reports the millivolts and stops there.

So an ORP electrode cannot be calibrated the way a pH electrode is. It gets checked against a known solution, then accepted or replaced. ZoBell's is the usual one, commonly quoted at +228 to +231 mV against an Ag/AgCl reference at 25 °C, roughly +428 mV on the standard hydrogen electrode scale. The exact value shifts with the formulation and with the reference fill, and ZoBell's is strongly temperature dependent. Take the number from the certificate for the standard on the shelf and from the manufacturer's temperature table. Not from memory.

ORP is reference-dependent as well. At 25 °C a saturated Ag/AgCl reference sits 197 mV below SHE, a 3.5 M KCl reference 205 mV below, and a 3 M KCl reference 210 mV below. Same solution, different numbers, depending on the scale it is quoted against. Two ORP readings from two instruments are not comparable until the reference and its fill are known.

What ORP mV reading do I need for proper disinfection in a cooling tower?

There is no single right number. Published control bands for oxidizing biocide in cooling water run around 550 to 650 mV, and the setpoint means nothing until it is tied to a measured free chlorine residual on that system.

The 650 mV figure that gets quoted everywhere came out of 1970s drinking-water disinfection literature, where a redox potential near that value was associated with near-instant virus inactivation. It survives today in state pool codes and vendor literature. It is not a current requirement anywhere that matters here.

ORP does not appear as a guideline value in current WHO drinking-water guidance — it shows up there as an indicator of how well chlorine is working, not as a limit. ASHRAE Standard 188 does not set one either; it requires a written, site-specific water management program rather than a water chemistry number. In cooling water specifically, published setpoints for oxidizing biocide control are lower than the drinking-water figure, commonly given as +550 to +650 mV. And every source that gives a band says the same thing about it: determine the setpoint for the individual site.

ORP is not a chlorine measurement. It responds to oxidizing power, and pH moves it hard. The HOCl/OCl− split has a pKa near 7.5 at 25 °C. Free chlorine is about 78% HOCl — the strong oxidizer — at pH 7.0, about 26% at pH 8.0, and about 5% at pH 8.8. Most towers run in that pH 8.0 to 8.8 band, where the same 1 ppm free chlorine reads far lower in millivolts than it would at pH 7.

Set the number empirically. Run DPD free chlorine tests at several ORP readings and pick the mV that lands on the target residual — continuous residuals around 0.5 ppm free chlorine are commonly cited, with periodic shock well above that. Then build that correlation again whenever pH, cycles of concentration, or the biocide changes. Legionella risk management belongs to the written water management program and to a water treatment specialist. Not to a setpoint on a controller.

Why is my dissolved oxygen probe reading low and responding slowly, and is optical better than galvanic or polarographic?

A DO probe that reads low and responds slowly is nearly always a fouled or damaged membrane on a galvanic or polarographic sensor. For dirty continuous service, optical is the better choice: no membrane, no flow requirement.

Membrane sensors consume oxygen at the cathode. They need flow across the face. A probe sitting in dead water depletes its own sample, reads low, and keeps sliding down. There is no universal minimum velocity — published figures span from a few millimetres per second on designs optimised for low flow up to tens of centimetres per second on older cells — so take the number off the datasheet for the sensor in hand rather than a rule of thumb.

Check in order: membrane torn, wrinkled, or coated; electrolyte depleted or holding a bubble; anode blackened or fouled; then flow past the tip. A polarographic sensor also needs polarization time after power-up, commonly quoted as 5 to 15 minutes and running to hours on some models and after a membrane and electrolyte change. A low reading right after a power cycle may be nothing worse than a cold start.

• Optical (luminescent): no membrane, no electrolyte, no warm-up, no flow requirement, and it does not consume oxygen. A completely stagnant sample still has to be exchanged to be representative of the process, but the sensor itself is not depleting it. The sensing cap is a consumable — manufacturers state 1 year on some models and 2 years on others. Usual choice for aeration basins and any fouling service.
• Galvanic: self-polarizing, reads immediately with no warm-up, but the anode is sacrificial and consumes itself even with the instrument powered off, which limits shelf life. Still needs flow.
• Polarographic (Clark): needs an applied bias, a warm-up, and flow, but amperometric sensors of this type remain the common choice for trace ppb DO in boiler feedwater and condensate. Optical trace sensors now reach the low ppb range as well, so check the stated resolution against the specification rather than assuming.

What is the difference between a bump test and a calibration on a gas detector?

A bump test proves the sensor and the alarms respond to gas. A calibration adjusts the reading so the number is correct.

Bump test: apply a known gas at a concentration and for an exposure time sufficient to activate every alarm setting, confirm the instrument alarms and the reading responds, adjust nothing. Calibration: zero on clean air or a zero-air cylinder, apply a certified span gas, adjust the reading to match the cylinder value.

OSHA's 2013 bulletin on direct-reading portable gas monitors carries the ISEA position on this: a bump test or calibration check before each day's use, in accordance with the manufacturer's instructions. An instrument that fails either one goes to full calibration before it goes back in service. Neither OSHA nor ISEA puts a stopwatch on the gas. The requirement is enough concentration and enough time to trip every alarm.

The trap is reading a passed bump test as an accurate instrument. A sensor can alarm reliably and still read badly low on span, which matters the moment someone uses the number instead of the alarm. Full calibration follows the manufacturer's stated interval; where an employer leans on less frequent verification, the ISEA position is that the interval between checks should never exceed 30 days. Record the as-found readings either way.

What causes false H2S alarms on a 4-gas monitor, and how long do electrochemical H2S and CO sensors last?

Most false H2S alarms are cross-interference from another gas, or a cell at the end of its life. CO and H2S cells typically last 2 to 3 years in service.

Check cross-interference first. The numbers belong to the specific cell, and published tables differ so much between manufacturers that quoting a general one does more harm than good. Pull the table for the sensor in the instrument. Set two current published tables side by side and you can see how wide the spread is, and what little does hold across both:

• Phosphine reads strongly positive — 55% on one table, 80% on the other.
• Nitrogen dioxide reads negative, around -21% to -25%. That is the dangerous direction: it suppresses a real H2S reading rather than creating a false one.
• Hydrogen reads essentially nothing — 0% and 0.08% on the two tables. This retires an old rule of thumb. Current H2S cells are filtered or electronically compensated for hydrogen, and an old table does not describe a new sensor.
• Everything else — sulfur dioxide, hydrogen cyanide, hydrogen chloride, chlorine, ammonia, nitric oxide — runs from zero to enormous depending on whose cell it is, and the two tables disagree even on the sign for some of them.

Every one of these tables carries the same warning: the figures apply to new sensors and shift as the cell ages.

Non-gas causes move an electrochemical cell too. Rapid humidity or temperature swings, water on the sensor filter, a hit of solvent or aerosol — brake cleaner, penetrating oil, hand sanitizer.

On life: most CO and H2S cells run 2 to 3 years in service and are commonly warranted 24 months. Unpowered spares age on the shelf as well, so check the manufacturer's stated storage window before stockpiling a drawer full. Replace on evidence rather than the calendar. Climbing span adjustment at each calibration, slower response to bump gas, an unstable zero: the cell is finished.

Catalytic bead vs infrared LEL sensor - which one for methane detection?

For methane in normal air, infrared is the better sensor. Use catalytic bead when hydrogen has to be detected, or when the mix of combustibles is broad and unknown.

• Infrared (NDIR) reads hydrocarbon C–H bond absorption, so it works with no oxygen present and cannot be poisoned. It is blind to hydrogen, which has no C–H bond, and a hydrocarbon-band sensor does not respond to non-hydrocarbon combustibles such as ammonia or carbon disulfide.
• Catalytic bead oxidizes gas on a heated bead and therefore needs oxygen: the reaction wants roughly 12% O2 to run properly, and below about 10% O2 the sensor cannot be relied on to read to full scale. It can therefore read low or zero in an inerted space that is full of gas.
• Poisons — silicones above all, plus sulfur compounds, lead, and halogenated hydrocarbons — permanently kill a catalytic bead, and the instrument may still zero normally and look healthy afterward.
• Over-range. A bead driven well above 100% LEL can saturate and read back down through scale. IR does not.

Either way, the calibration gas has to match the application, or the correlation factor has to be applied. A sensor spanned on methane and used on pentane vapor does not report correct %LEL. And if the detector sits inside a SIL-rated safety instrumented function, the sensor technology, voting arrangement, and proof-test interval belong to the SIS design under IEC 61511, not to a general preference between the two.

Toroidal vs contacting conductivity sensor, and how do I choose a cell constant of 0.01, 1.0, or 10?

Contacting on low-conductivity water, toroidal on higher conductivity or anywhere fouling is the problem. Match the cell constant to the range: 0.01 for ultrapure, 0.1 for pure water, 1.0 for general process, 10 for concentrated solutions.

A contacting cell puts electrodes in the fluid. A toroidal — inductive, electrodeless — cell uses two coils and wets no metal at all, which is what lets it tolerate coating, fouling, and corrosive service. The trade is sensitivity at the bottom of the scale. A typical toroidal sensor is specified from roughly 50 µS/cm up to 2,000,000 µS/cm on a single cell factor, which rules it out for condensate and boiler feedwater. Some models claim lower. Check the specification, not the category.

• K = 0.01 cm⁻¹: ultrapure water, from 0.055 µS/cm (18.2 MΩ·cm) up to a few µS/cm.
• K = 0.1 cm⁻¹: pure water and condensate, roughly 1 to 100 µS/cm.
• K = 1.0 cm⁻¹: general process and potable water, roughly 10 to 2,000 µS/cm.
• K = 10 cm⁻¹: brines, acids, caustic — 1,000 µS/cm and up into the hundreds of mS/cm.

Those ranges are typical rather than specifications, and manufacturers overlap them heavily. The set of four is standard across the industry, and between them they cover essentially the whole conductivity range.

Most conductivity arguments actually end at temperature compensation. A dilute solution changes about 2% per °C, so every reading has to be normalized to 25 °C. Pick the wrong compensation curve — neutral salt versus high-purity versus cation conductivity — and it is worth many percent on the same water.

How do you convert uS/cm to ppm TDS?

Multiply µS/cm by a factor between 0.5 and 0.7 — 0.5 on the NaCl scale, about 0.7 on the 442 scale — and treat the result as an estimate, not a measurement.

1,000 µS/cm is 500 ppm on the NaCl (500) scale and 700 ppm on the 442 (700) scale. A 40% spread on identical water. That is why two handhelds disagree, and why the first question is which scale each one is set to rather than which one is right. KCl-referenced instruments land around 0.50 to 0.55.

Conductivity only sees ions. Non-ionic dissolved solids — sugars, most organics, dissolved silica — contribute nothing to conductivity and everything to real TDS, so the estimate reads low on water carrying a genuine dissolved load. And when a permit or a product spec says TDS, it usually means the gravimetric lab test: filter, evaporate the filtrate, and dry to constant weight at 180 °C (Standard Methods 2540 C, EPA Method 160.1). A conductivity meter is not a substitute for it.

What is the minimum conductivity in uS/cm for a magnetic flow meter to work?

5 µS/cm is the usual general-purpose minimum. The figure is model- and bore-specific, though — published tables run from about 1 µS/cm to 5 µS/cm on the same meter depending on line size.

Read the specification for the tag in front of you rather than a general number. Most general-purpose magmeters state 5 µS/cm, but the published tables vary by bore on the same model. Yokogawa's AXF, for one, states 5 µS/cm or greater for 2.5 to 10 mm, 1 µS/cm or greater for 15 to 125 mm, and 3 µS/cm or greater for 150 to 400 mm, and recommends its enhanced dual-frequency option below 3 µS/cm. Small bores generally need more conductivity than mid-size bores on the same model, since electrode area and generated signal both shrink with diameter. The pattern is not monotonic across the whole size range, which is exactly why the table matters.

Capacitance-electrode designs built for low-conductivity service go lower still. Demineralized water at 0.055 to 1 µS/cm falls below almost every magmeter. Hydrocarbons — crude, gasoline, diesel, solvents — carry essentially no free ions, and a magmeter will never work on them at any size or excitation frequency.

Near the limit the failure is a noisy, jumpy reading rather than a dead one, and the fixes are electrical: solid process grounding through grounding rings or a ground electrode, shielded signal cable inside the manufacturer's stated length limit, enhanced or dual-frequency excitation. The pipe also has to be full. The common installation rule is 5 pipe diameters straight upstream and 2 downstream, and a partly empty pipe reads low no matter how conductive the fluid is. Some meters are specified to hold accuracy with less straight run than that — a datasheet claim to check, not to assume.

Why is my turbidity analyzer reading high when the water looks clear?

At low turbidity the reading is usually air bubbles or stray light rather than particles. The eye cannot resolve turbidity below about 5 NTU in any case, so clear-looking water says nothing about whether 0.3 NTU or 2 NTU is correct.

• Entrained air. Bubbles scatter light exactly like particles. Sample lines taken off a pump discharge or through a throttled valve that flashes will make them; a bubble trap, a relocated tap, or standing a grab sample 2 to 5 minutes will tell which it is.
• Fouled or scratched optics. Film or scratches on the window or flow cell read as turbidity and drift upward over weeks.
• Stray light. An aging lamp, a cracked cell, dust inside the optics, or light leaking into the chamber — the dominant error source below 1 NTU.
• Condensation on a cell holding sample colder than ambient.
• Biofilm in the sample line, which is a true reading of a real problem in the sample line and not in the process.

Prove it with standards before touching the process. Read a sealed secondary standard, then verify against a primary standard — freshly prepared formazin, or a stabilized formazin standard within its date. If the standards read correctly and the process still reads high, the problem is upstream of the optics.

What is the difference between NTU and FNU turbidity units?

NTU is white-light scatter under EPA 180.1. FNU is 860 nm infrared scatter under ISO 7027. The two do not produce the same number on the same water.

Both measure light scattered at 90° from the incident beam, and both are calibrated with formazin. The light source is the difference. EPA 180.1 uses a tungsten lamp at a color temperature of 2200 to 3000 K, with the detector at 90° ± 30° to the incident beam; ISO 7027 uses a narrow-band LED at 860 nm ± 30 nm, with the detector at 90° ± 2.5°.

Scattering depends on particle size relative to wavelength, so the two do not agree on the same water, and how far apart they land depends on what is in it. There is no valid conversion factor. Do not apply one. EPA 180.1 is written for 0 to 40 NTU directly; above that the sample is diluted with turbidity-free water and the result multiplied back by the dilution factor.

Where it bites is color. Tannin-stained or otherwise colored water absorbs visible light and reads artificially high in NTU. The infrared FNU measurement largely ignores it, which is why the infrared method is favored for surface water carrying dissolved organics. For US drinking water compliance, report in the unit the approved method produces.

What causes retention time shift on a process gas chromatograph?

Carrier gas pressure or flow first, column temperature second, leaks third — in that order.

If every peak moves the same direction, the whole chromatogram is sliding, and the cause is carrier flow or oven temperature. Everything peaks late: flow is down — leak upstream, restriction, plugged frit, drifting regulator. Everything peaks early: flow is up, or the column is running hot. Compare actual column head pressure against the method value before changing anything else, and leak-check with an electronic detector rather than soap.

Ambient temperature gets overlooked on field-mounted process GCs. A shelter with weak heat or failed HVAC swings the oven and the pressure regulator with the weather, and the shift then tracks the day rather than the sample. If only some peaks move, or only one valve group moves, look at the column instead — phase loss, active sites, contaminated guard column — or at valve timing. Method work on a custody-transfer or emissions-reporting GC belongs to the analyzer specialist and the method owner. Column selection, valve timing, revalidation: not a field adjustment.

Which way should impulse lines slope for gas, liquid, and steam service?

Gas slopes up to a transmitter above the tap. Liquid and steam slope down to a transmitter below the tap. The minimum slope is 1 inch per foot in every case.

• Gas: tap on top of the line, transmitter above, legs sloping continuously up toward the process at 1 in/ft (about 83 mm/m) so condensate drains back to the process and never collects in a leg.
• Liquid: tap on the side of the line, transmitter below, legs sloping continuously down toward the transmitter so gas bubbles rise back to the process. Trapped gas in a liquid leg is the classic cause of a DP that wanders and will not hold zero.
• Steam: tap on the side, condensate seal pots at the same elevation on both legs, transmitter below, legs sloping down and both filled to the same height with condensate.

Both legs must also run at the same temperature and the same slope. Two legs at different temperatures hold different fluid densities, and the false differential that comes out of that will not zero away. 1 in/ft is a minimum, not a target — some references call for a steeper 1 in 10. Keep the runs short. Every extra foot of impulse line adds response time and one more place for gas or condensate to collect. A zero shift on a steam DP after an outage almost always means a leg boiled dry: refill it and re-zero.

When this page runs out

This page covers what a technician can settle at the panel with buffers, standards, and a meter. It stops short of the rest on purpose. If the job needs an accredited calibration certificate, that is an accredited calibration laboratory, and Bridges Industrial will say so. Legionella risk management in a cooling system belongs to a water treatment specialist and the facility's written water management program, not to an ORP setpoint. Drinking water compliance reporting is governed by the approved method and the state primacy agency. Process GC method development belongs to the analyzer specialist, and any gas detector inside a SIL-rated safety instrumented function belongs to the SIS design.

And an electrode that has been cleaned, rehydrated, and calibrated against fresh buffer and still slopes below 85 to 90 percent is not a troubleshooting problem at all. It is a consumable at the end of its life. Buy a new one.