A leak detection and repair programme is an inventory with a measuring stick: thousands of valves, flanges, pump seals and connectors, each visited, each read, the leakers tagged and repaired and re-read. The programme's credibility rests on the instrument in the surveyor's hand, and the FID-versus-PID question is really a question about which parts of the inventory each instrument should read. This article divides the work the way experienced LDAR coordinators do.
The shape of a survey
Components are read at their leak interface — the stem packing, the flange face, the seal gland — with the probe held where escaping vapour concentrates. Readings above the programme's leak definition trigger tagging, repair within the allowed window, and a verification re-read. Between components, the surveyor crosses open plant where fugitive clouds drift unsourced, and the instrument is a screening tool again.
Two measurement temperaments are therefore needed in one walk: defensible quantification at the interface, and fast sensitive screening between interfaces. Programmes that force one instrument to fake the other temperament produce numbers nobody can defend and maps nobody can use.
FID at the interface
Flame ionisation is the reference temperament: broad hydrocarbon response, quantitative behaviour and the method-standard pedigree that LDAR regulations are written around. At the component, the FID reading is the record — the number that decides repair and survives audit. Its hydrogen appetite and flame management are the price, softened in modern portables by automatic ignition, flame-temperature monitoring and fuel-state displays, as on the PTM100.
The PTM200 extends the FID family to the regulatory aggregate: direct methane, total hydrocarbon and NMHC readings for the emission numbers that sit above the component inventory in the same permit universe.
PID between the components
Photoionisation is the screening temperament: instant start, no fuel, and low-ppm sensitivity to many solvents and aromatics let a surveyor sweep tank farms, loading racks and remediation grids quickly, marking where the vapour geography concentrates. Its compound-relative response disqualifies it as the interface referee but makes it the best scout in the programme.
The dual-detector instrument — PTM100 pattern, FID and PID sharing one sample path and one data trail — dissolves the scheduling problem: screen with PID on the walk, quantify with FID at the component, one surveyor, one record. For programmes large enough to matter, that combination is less a luxury than the arithmetic of surveyor-hours.
Records close the loop
An LDAR reading without a component identity is weather. The survey's data structure — component tag, reading, detector, calibration state — is designed before the first walk, and the instrument's library and logging serve it: the PTM100's built-in database of more than 200 VOC gases exists so that response knowledge travels with the reading. Programmes are audited on their trails, not their intentions; equip the trail.
Anatomy of a component survey
A survey day is a list before it is a walk: the component inventory — valves, flanges, pump seals, compressor seals, connectors, open-ended lines, sampling connections — each with an identifier that ties the reading to the iron. The surveyor works the list with the probe at each interface, dwelling long enough for the instrument to respond and the reading to settle, recording the peak against the component ID. Method discipline lives in those dwell times and probe placements; the difference between a survey and a stroll is that every number can be traced to a component, a time and an instrument state. Programmes that skip the inventory step generate weather reports about the plant's general leakiness; programmes that keep it generate a repair list.
FID versus PID at the component face
The flame ionisation detector burns the sample in a hydrogen flame and counts carbon — nearly every organic vapour registers, methane and ethane included, with response factors close to unity across the hydrocarbon family. That completeness is why emission regulations built the standard method on FID: nothing organic hides from it. Its costs are the hydrogen supply, the flame's behaviour in wind and oxygen-poor pockets, and an instrument that demands a competent operator. The photoionisation detector trades completeness for convenience: no flame, no hydrogen, instant readiness — but a census limited to compounds its lamp can ionise, reported in equivalents of its calibration gas, blind to methane entirely. A survey that must see methane with a PID is over before it starts; a survey of solvent valves with an FID is using a cannon correctly but expensively. The component list decides.
Response factors and the honest reading
No survey reading is raw truth; it is the instrument's response to that compound, translated. FID response factors vary modestly — oxygenated and halogenated compounds respond less per carbon — and PID correction factors vary widely, from near unity for aromatics to several-fold for light alkanes the lamp barely touches. Programmes publish their factor tables and apply them before comparing readings against repair thresholds, because a 400 ppm reading that corrects to 1,600 ppm is a different repair decision. Surveyors should carry the factors for their plant's actual compounds, not the instrument's default assumptions.
Thresholds, leaks and the repair loop
Programmes define a leak in ppm at the interface — figures like 500 or 2,000 ppm appear in various regimes, with first repair attempts inside days and final repair inside weeks, and delay-of-repair provisions for components that cannot be freed without a shutdown. The details belong to each site's permit, but the operational shape is universal: survey, tag, attempt, re-survey, document. The re-survey after repair is the step that separates programmes from performances, because a tightened flange that still reads 3,000 ppm is not a repair; it is a rumour of one. Records close the loop — component, reading before, action, reading after — and auditors read exactly that chain.
From components to fugitive totals
Multiplying average emission factors by component counts gives inventories; multiplying measured concentrations by correlation equations gives better ones; enclosing and measuring gives the best and costs the most. However the arithmetic runs, the survey data is the input, and its quality caps the inventory's credibility. Organisations reporting fugitive emissions upward — to regulators, to parent companies, to ESG disclosures — should treat survey discipline as financial discipline: the numbers will be audited like money, because increasingly they are money, in carbon prices, product losses and licence to operate.
People: the surveyor is the instrument's other half
Two surveyors with the same instrument produce different surveys: pace, probe placement, dwell, honesty about re-checking a suspicious reading. Training and occasional accompanied audits — a second surveyor re-walking a sample of components — keep the human half calibrated. Programmes should also protect surveyors from their own findings' pressure: a surveyor rewarded for low counts will eventually produce them. Reward completed lists, traceable records and found leaks verified by re-survey — the behaviours that make the data true.
The inventory is the programme
Everything in LDAR descends from the component inventory, and inventories decay: plants modify, blind, reroute and demolish, while the database remembers a facility that no longer exists. Mature programmes treat the inventory as a living register with its own maintenance — field verification on a cycle, modification workflows that add and retire components at the project gate, tagging that survives paint and weather. The surveyor's route is only ever as good as the register's truth, and a programme that finds 'components' that are actually pipe stubs, or walks past untagged valves, is measuring its paperwork. Budget the register's upkeep as a line item; it is the cheapest emissions integrity money can buy, and its absence is the most common finding when programmes are audited by people who walk the plant.
Background, wind and the surveyor's patience
Field readings breathe with the plant's background: a surveyor downwind of a tank vent reads the facility, not the flange. Method discipline handles it — background readings taken and recorded, approaches made from upwind where the layout allows, dwell times that let a local peak declare itself against the background's wander, and the honesty to mark a component 'inaccessible to a valid reading today' rather than harvest a number from contaminated air. Programmes that punish such honesty get numbers; programmes that reward it get data. The distinction shows up years later, when the inventory's emission estimates are compared against a fence-line measurement and only one kind of programme is unsurprised.
Optical imaging joins the kit
Optical gas imaging cameras have rearranged survey economics: a camera sweep finds the big emitters across a unit in minutes — the open thief hatch, the failed seal, the venting compressor — that a component-by-component sniffer survey would take days to reach. The mature workflow is therefore two-tiered: imaging for discovery and prioritisation, FID or PID at the component face for the quantified, recordable reading the programme's thresholds are written in. Imaging does not retire the sniffer; it promotes it, from search party to measuring instrument. Programmes adopting cameras should rewrite their procedures around the tiers explicitly, because a camera finding without a quantified follow-up is a photograph, and a sniffer survey without imaging is a census taken in the dark.
People, pace and the economics of the found leak
Close the economics: a surveyor-day costs what it costs, and its value is the repairs it causes. Programmes therefore track yield — leaks found per hundred components, mass avoided per surveyor-day, repair closure time — and route effort toward the component classes and units with the richest history. This is not corner-cutting; it is the same triage medicine practises, and it depends entirely on the records being honest, because triage on flattered data sends the surveyor to the healthy wards. The found leak, verified closed by re-survey, is the programme's unit of virtue; everything in this long article — principles, factors, inventories, imaging — exists to manufacture that unit reliably.
Contractor surveys and the client's duty
Much LDAR is performed by contractors, and the arrangement has a known failure shape: the client buys readings, the contractor sells readings, and neither owns the truth between them. Healthy contracts specify the method's bones — inventory verification duties, background and dwell discipline, factor tables to be used, re-survey percentages, data formats with audit trails — and they pay for completed, verified component lists rather than for low leak counts. Client-side auditing rides along: accompanied re-walks of a sample, blind re-measurement of tagged components, records sampled against the plant's physical reality. None of this expresses distrust; it expresses the understanding that emissions data, like accounts, is a representation that both parties will one day have to defend together.
Small plants and the proportionate programme
LDAR's literature is written by refineries, but its logic scales down: a chemical store, a solvent plant, a biogas upgrading skid each has a component census in the hundreds rather than the hundreds of thousands, and a programme of quarterly walks with a PID, a repair notebook and a re-check habit captures most of the value at a fiftieth of the bureaucracy. Proportionality is a virtue, not a compromise — the programme that a small plant will actually run for a decade beats the refinery-grade regime it would abandon by spring. What must not scale down is the honesty: factors applied, backgrounds noted, repairs verified. Small programmes can borrow the big programmes' conscience without borrowing their paperwork.
The leak that pays for the year
Every programme has its folklore: the passing valve found at 40,000 ppm that had been venting product worth more than the survey contract; the compressor seal whose repair ended a neighbourhood odour saga; the open-ended line, capped in an afternoon, that had been the plant's largest single source for years. The folklore is not decoration — it is how programmes survive budget rounds, by reminding everyone that fugitive emissions are product, money and licence escaping through the same holes. The instruments in this article's title are, in that light, recovery tools: the FID and the PID earn their keep not in the readings they produce but in the leaks that stop existing because the readings did.
Repair verification and the second reading
The repair loop's integrity rests on the second reading: the re-survey after the wrench, performed by the method, recorded against the component, without which 'repaired' is a hope wearing a tag. Programmes should protect the second reading's independence — a different surveyor where the fleet allows, the same instrument class and factors where it must — and should treat a failed verification as information about the repair method, not merely about the component: the packing that never holds, the gasket face that is scored, the valve whose design leaks by intention of its clearances. Those patterns, accumulated, redirect capital toward the components that cannot be maintained into tightness, which is where LDAR data quietly becomes engineering.
LDAR's place in the carbon account
Methane's climate weight has promoted LDAR from compliance chore to carbon instrument: the same surveys that satisfy the permit now feed the inventory, the ESG report and, increasingly, a price. The promotion raises the evidential stakes — inventories are audited, compared against satellite and fence-line estimates, and defended in public — and it rewards exactly the virtues this article has catalogued: complete inventories, honest factors, verified repairs, weather-aware surveys. Programmes built carelessly for a forgiving regulator will not survive the carbon auditor; programmes built honestly for the physics will not notice the transition. Build for the physics.
Survey data's second life in design
Close the loop upstream: LDAR archives are design intelligence. The component classes that leak chronically in this plant's service — the valve packing that hates this temperature, the flange facing that will not hold this gasket — are specifications waiting to be changed in the next project, and the survey data is the only witness that can name them with numbers. Engineering teams should receive the LDAR league tables as a design input, the way they receive corrosion data, because a plant that keeps buying its leakiest components is paying for the same emissions twice: once in the survey and once in the product. The surveyor's walk thus ends, years later, in a purchasing specification — the longest and most valuable journey any reading in this article makes.
The surveyor's kit, itemised by a veteran
The veteran's kit tells the programme's story in objects: the instrument with its calibration fresh and its factor table laminated in the lid; the probe with its filter changed on rhythm rather than on failure; the wind sock's pocket substitute — a ribbon, a tissue, anything that confesses the air's direction; the tag roll and the camera; the water bottle that keeps the surveyor's pace honest in August. Programmes should kit their surveyors as deliberately as they calibrate their instruments, because the survey's quality is manufactured at the kit bench the night before as surely as at the flange face. The reading is the last step of a chain, and chains are audited at their weakest link, which is rarely the flame or the lamp.
Counting what counts
Programmes drown in counts — components walked, readings taken, hours spent — while starving for the one count that justifies them: emissions that no longer happen because a survey found their source and a repair closed it, verified. Keep that count at the top of the report, in mass, with its method stated, and let the activity counts follow as supporting detail. The inversion disciplines everything beneath it: routes bend toward the leaky parishes, repairs are verified because the headline number depends on them, and the programme's annual story becomes a slope descending, which is the only story an emissions programme exists to tell.
The survey's first hour
Programmes are decided in their first hour more often than their reports admit: the calibration witnessed rather than assumed, the background taken where the wind makes it honest, the first components dwelt on long enough to set the day's tempo, the surveyor's notebook opened with the weather written in it. Supervisors who ride the first hour occasionally — not to catch anyone, but to keep the standard a living thing rather than a laminated one — find the whole season's data improving behind the habit. Beginnings are leverage; the survey that starts honestly rarely needs to be rescued by its paperwork.
What the found leak teaches the plant
Each verified leak is a small biography of the plant's maintenance: the gasket material that this service eats, the valve whose packing schedule was written for a gentler temperature, the flange that was torqued by a procedure nobody follows anymore. LDAR programmes that feed those biographies back to the maintenance planners — not as scores but as stories with numbers attached — convert surveying from a compliance tax into the plant's own memory of its weak points. The plant that learns from its leaks buys fewer of them each year, which is the only trend in this entire subject that compounds in the owner's favour.
The programme's annual story, told straight
Write the programme's annual story in four sentences and refuse all others. We walked this many components and our register is this honest. We found this much mass escaping, at these classes of iron, and verified this fraction of it closed. Our readings were made by these principles, against these factors, in this weather, and a sample of them was independently re-walked with this agreement. Next year the effort moves here, because the data says the leaks live there. A programme that can tell that story owns its subject; a programme that needs forty pages owns a filing system. The instruments in the title are the story's verbs — everything else in this long article exists to make the sentences true.
The surveyor's creed, in four lines
Write the creed where the instruments are charged. I measure the component in front of me, in the weather that is given, and I write down what the instrument said — not what the plant needed said. My factors are the plant's own, my backgrounds are taken upwind, my dwell is long enough for the truth to arrive. A leak I find is a repair I cause; a leak I flatter is an emission I co-author. The creed takes thirty seconds to read and a career to keep, and programmes that post it are announcing, to their own people first, which of the two they intend to be.
Colophon: the walk continues
The surveyor's walk is the industry's longest-running measurement: it began with noses and soap solution, graduated through the sniffer's generations, and now proceeds with lasers overhead and cameras in hand — yet its essence has never changed, which is a person moving deliberately through a plant, paying instrument-grade attention to the places where the process meets the air. Honour the walk in the programme's design and the rest follows: honest numbers, verified repairs, an inventory that descends year by year. The plant is always leaking a little; civilisation is deciding how little, and proving it.