August 17, 2026

A waste heat recovery boiler sitting idle because it was rushed into service — wrong foundation spec, bypass damper leaking, steam drum not properly vented during first firing — is not a hypothetical. It happens, and when it does, the upstream process usually has to throttle back or shut down entirely while the boiler is pulled apart and corrected. That lost production time typically runs weeks, and the rework cost on a mid-size installation can easily absorb a significant chunk of the original equipment budget. Getting the installation sequence right the first time is not a quality aspiration; it’s a financial necessity.

Installing a waste heat recovery boiler involves six core phases: site preparation and foundation work, equipment receiving and rigging, pressure-part assembly and flue gas ductwork connection, auxiliary system integration (feedwater, blowdown, instrumentation), refractory curing and steam blowing, and final commissioning with performance verification against the design heat-recovery targets.

What makes WHRB installation genuinely different from a conventional fired boiler job is the dependency chain running in both directions — the upstream process dictates flue gas conditions the boiler must tolerate, while the boiler’s readiness dictates when that process can run at full capacity again. Compress the schedule in the wrong place, and you do not save time; you create a bottleneck that costs more to fix than the weeks you thought you were saving. Understanding exactly what each step demands — and where the real risk concentrations are — is what separates a clean startup from an expensive one.

Pre-Installation Engineering: Site Survey, Process Gas Analysis, and Equipment Matching

Before a single structural steel anchor gets welded or a single pressure part gets lifted into position, the real work of a WHRB project happens at a desk — and on the process floor with a clipboard, a pitot tube, and sometimes a dust sampling train. Get this phase wrong and you end up with a boiler that either underperforms chronically, corrodes prematurely, or trips the upstream process every time conditions drift outside a narrow band nobody documented properly. It’s the most unglamorous part of the project and the one most likely to get compressed when schedules tighten.

Flue Gas Characterization: The Non-Negotiable Starting Point

Every WHRB is essentially a custom heat exchanger designed around one specific gas stream. Treating the design input as a rough estimate is how you end up with a unit that’s perpetually undersized or one that corrodes through its first economizer tubes in under two years.

The gas audit needs to capture volumetric flow rate in Nm³/h — not just at rated process load, but at turndown and at peak — because WHRBs servicing cement kilns or electric arc furnaces see wide swings in both flow and temperature across a single shift. Inlet temperature is the headline figure (commonly 350°C to over 1,000°C depending on whether you’re downstream of a gas turbine, a glass tank furnace, or a rotary kiln), but it’s the combination of temperature, dust loading, and gas chemistry that actually drives design choices.

Dust loading deserves particular attention. A gas stream from a cement preheater exit carrying 40–80 g/Nm³ of alkaline dust behaves nothing like the relatively clean exhaust from a reciprocating engine genset. High dust loading pushes you toward wider tube pitches, extended soot-blowing provisions, and possibly a pre-dedusting cyclone upstream of the boiler — all of which need to be sized and priced before procurement, not retrofitted after the boiler is running and plugging every 96 hours.

Gas composition — SO₂, HCl, NOx, and water vapor content — determines your acid dew point, which in turn sets a hard floor on the minimum metal temperature you can allow in the economizer. Ignore this and you will see cold-end corrosion within months, usually most aggressively at the economizer outlet headers where local gas velocities drop. For waste streams containing meaningful HCl concentrations (municipal solid waste incinerators, certain chemical process exhausts), the dew point can sit above 130°C, which has direct implications for feedwater inlet temperature and economizer tube material selection.

Heat Balance Calculations and Section Sizing

Once the gas-side data is solid, the heat balance tells you what the boiler can actually deliver: evaporation capacity in t/h, required heating surface area for each section (superheater, evaporator, economizer, and air preheater if included), and the steam parameters — say, 2.5 MPa saturated, or 4.0 MPa / 380°C superheated, depending on the downstream steam system the plant operates. A well-matched design can recover 60–85% of the available exhaust enthalpy; the actual figure depends on the pinch point temperature difference you’re willing to accept and the minimum stack temperature your acid dew point permits.

The pinch point typically runs 15–35°C in a well-designed unit. Narrow it aggressively to chase recovery efficiency and you shrink margin against process fluctuations — one short-duration temperature spike from the upstream furnace and the evaporator tubes go into departure from nucleate boiling territory faster than any operator can respond manually.

Natural circulation works for most industrial WHRBs in the 1.0–6.3 MPa range and keeps the system mechanically simple, which matters for long-term maintainability. Forced circulation becomes necessary at lower steam pressures where density difference between riser and downcomer water is insufficient, or in horizontal gas-flow configurations where the circulation geometry doesn’t favor gravity-driven flow. It’s not inherently worse — it just adds a circulation pump, a pump seal maintenance schedule, and a failure mode that needs to be covered in the plant’s protection logic.

Boiler Configuration and Bypass System Selection

The choice between horizontal and vertical gas-flow arrangement is partly dictated by available plot space and partly by dust behavior. Vertical upward-flow units tend to self-clean better in high-dust applications because particles don’t settle on tube banks the same way they do in a horizontal arrangement. That said, vertical units are taller and can run into headroom constraints in existing buildings — a ceiling-height check sounds trivially obvious but it genuinely catches people out, especially in brownfield retrofits where the upstream process equipment was never designed with a boiler in the headspace.

A bypass damper system isn’t optional if the upstream process runs independently of the boiler — and for most industrial WHRBs, it does. Cement plants don’t stop the kiln because the boiler needs a tube inspection. The bypass flue needs to be sized for full gas flow, the damper actuator needs to be fast enough to protect against temperature spikes (pneumatic is generally faster and more reliable in a dusty environment than electric), and the thermal expansion of the connecting ductwork between process and boiler needs to be absorbed by expansion joints designed for the actual temperature cycling range, not a generic catalog selection.

Civil and Structural Validation

Foundation bearing capacity — usually in the range of 150–350 kN/m² for mid-size WHRBs, depending on total operating weight including water — needs to be confirmed by a geotechnical assessment, not estimated from a similar boiler on a different site. Steam drum elevation drives the required headroom, and the horizontal distance from the process gas outlet to the boiler inlet affects both duct heat loss and thermal expansion management. Keep that connection duct short where the layout allows; every unnecessary meter adds surface area for heat loss and another potential leak point.

Code Compliance and Manufacturer Qualification

This is the checkpoint that EPC clients in overseas projects skip at their peril. Before issuing a purchase order, confirm that the manufacturer holds a valid pressure vessel fabrication license compliant with the applicable code for the destination country — ASME Section I for North American projects, EN 12952 for European Union markets, GB/T 16507 for China-origin units being accepted under Chinese regulatory frameworks, or the local equivalent. A boiler that arrives without the right design certificate will be held up at customs or by the local boiler inspectorate, and the cost of delay on a running project dwarfs whatever was saved on the unit price.

Ask for the manufacturer’s most recent third-party inspection records, welding procedure qualifications, and material mill certificates as part of the technical bid evaluation — not as an afterthought during FAT. Rushing this step is one of the more expensive mistakes a procurement manager can make on a capital project.

Foundation and Structural Steel Erection for WHRB Support Frames

The foundation is where a significant number of WHRB projects go quietly wrong. Not during commissioning — during year two or three, when a header develops a hairline crack, or a tube bank starts showing fatigue failures that nobody can immediately explain. Trace it back and you usually find a bearing point installed in reverse, or base plates that were grouted in a hurry and have since settled unevenly. Getting the civil and structural work right is unglamorous, but it is genuinely the difference between a unit that runs 20 years with predictable maintenance and one that bleeds downtime.

Concrete Foundation Preparation

Design drawings will specify compressive strength — C30 is the practical minimum for most medium-capacity WHRBs, and C35 or higher is common for units above roughly 30 t/h or where seismic loading governs. The embedded anchor bolts deserve more attention than they typically get on busy sites. Positional tolerance is ±2 mm from the design centerline; anything looser than that and you are already imposing a lateral offset into the base plate before the structure carries a single kilogram of load. On a boiler with a flue gas inlet at 600–900°C, the cumulative thermal stresses are high enough that you cannot afford to start with a pre-loaded condition.

Cure time is non-negotiable. Full design strength requires 28 days under normal curing conditions — longer if ambient temperatures drop below around 10°C, and you should be adjusting your schedule accordingly if the pour happens in late autumn. Placing heavy structural steel on concrete that has only reached 60–70% of its design strength is a shortcut that has caused measurable long-term settlement on more than a few projects. Mark the pour date on the drawings, post it on site, and do not let schedule pressure override it.

Erecting the Boiler Support Steel Frame

Column verticality needs to be held to ≤1 mm per meter of height, verified by a calibrated optical plumb or total station — not a bubble level. Beam levelness matters just as much; the steam drum and headers will not forgive a frame that slopes a few millimeters over its length, because gravity-assisted drainage and steam separation both depend on correct orientation. Diagonal bracing between columns is not optional, even if it looks over-engineered on paper. Duct connection flanges on the inlet and outlet are significant vibration sources, particularly on units handling exhaust from EAFs or cement kilns where gas flow is pulsed or turbulent. Omitting anti-vibration gussets at those flange connections is one of the more common — and more expensive — structural errors I have seen.

Sliding and Fixed Bearing Points

A medium-capacity WHRB (say, 15–50 t/h) will expand longitudinally by roughly 15–40 mm and laterally by 5–15 mm as it heats from ambient to operating temperature, depending on overall length and the specific inlet gas temperature. The fixed bearing point — the anchor — belongs at the steam drum end. Sliding pads or roller assemblies go at the opposite end and along expansion legs. Get this reversed and you are constraining thermal growth at the wrong location, which loads the headers and tube-to-header joints in a way the stress analysis never accounted for.

Independent Structural Review Before Any Lift

Before any boiler module is lifted into position, an engineer who was not part of the original erection crew should walk the frame and verify the load envelope. The combined loading case — dead weight of the boiler body, water-filled hydrostatic test load (which adds substantial mass over the operating steam load), plus applicable wind and seismic factors from the site’s design basis — must sit within the frame’s rated capacity with a margin. This is not a paperwork formality. A hydrostatic test on a large WHRB can add 40–80 tonnes of water weight that is not always accounted for in rushed site schedules.

Base plate grouting deserves a specific call-out: non-shrink grout, properly mixed, with adequate cure time before load transfer. Under-grouted or poorly bonded base plates allow micro-movement under thermal cycling. Over years, that movement translates directly into fatigue stress in the lower pressure part connections. It is a slow failure mode, which makes it especially dangerous because it rarely triggers an alarm until something actually cracks.

Boiler Module Lifting, Positioning, and Mechanical Assembly Sequence

Getting a large WHRB from the laydown yard to its final operating position is where a lot of projects quietly go wrong. The engineering drawings look clean. The crane shows up on time. Then someone decides to improvise the rigging angle on a 38-tonne tube bank section, and you spend two days straightening bent fin tubes and re-inspecting welds. The lifting and assembly phase has very little tolerance for improvisation.

Planning the Lift Before the Crane Arrives

Every module needs a documented lift plan — not a rough sketch, an actual engineered lift plan with certified load calculations. For a mid-size WHRB in the 20–50 t/h range, individual module weights typically break down roughly like this: bare tube bank assemblies often run 8–25 tonnes depending on tube diameter and row count, steam drums on units above 15 t/h can easily reach 15–40 tonnes, and prefabricated header assemblies with attached legs may add another 4–12 tonnes each. These figures depend heavily on tube wall thickness, drum length, and whether the manufacturer has pre-assembled any insulation or casing prior to shipment.

Select crane capacity with a minimum 25% margin above the heaviest single lift — in practice I’d push that to 30–35% if the site has any overhead obstruction that forces an extended boom angle, because effective capacity drops fast as the boom extends. On confined industrial sites, a 100–160-tonne lattice boom crawler crane is common for medium WHRBs; larger units or sites with deep foundation pits may need tandem lifting with two cranes, which adds coordination complexity and requires a separate tandem lift procedure signed off by a licensed rigging engineer.

Lifting lugs must be certified to the relevant load rating and match the attachment points specified on the manufacturer’s lifting drawing — not the nearest convenient nozzle flange or structural beam. This is worth checking on every single module before the hook goes on. Slings, shackles, and spreader bars should all carry current inspection certificates.

Assembly Sequence: Work Bottom-Up, Access-Out

The sequence matters as much as any individual step. A logical order for most horizontal gas-path WHRBs runs as follows: set the lower headers and support legs first, leveling and grouting them to the foundation anchor bolts before any tube load goes on; lower the tube bank assemblies into position next, seating them onto the support structure and verifying alignment in both planes; then bring in the upper steam drum, connecting downcomers and risers once the drum is fully positioned and leveled. After the pressure-containing envelope is mechanically fit up, membrane wall or casing panels close in the assembly. Inlet and outlet gas ductwork connects last.

That sequence is not arbitrary. It preserves clear working access during the stages when you need it most.

Tube-to-Header Fit-Up: The Detail That Determines Weld Quality

All tube ends must be cleaned of mill scale, oil, and any moisture before fit-up begins. Wire brushing alone is often insufficient on tubes that have been sitting in outdoor storage through a wet season — light grinding to bright metal is more reliable. Root gap and alignment tolerance is typically ±0.5 mm; exceeding that produces incomplete root fusion that won’t show up on visual inspection but will show up eventually as a leak or cracking failure in service. Use fit-up fixtures rather than relying on the welder’s eye. On large tube banks with hundreds of roll-expanded or butt-welded joints, even a 1 mm systematic error in fixture setup compounds badly.

The Open-Access Window: Install Internals Now, Not Later

Once casing panels go on, internal access is essentially closed. Internal baffles, flow distribution guides, and soot blower lances must be installed during the open assembly window — ideally after tube bank positioning but before membrane wall panels are welded in. Retrofitting a misaligned soot blower lance through a finished casing is a half-day job minimum, assuming you can even get the geometry right. Skipping a baffle because it “seems like a minor detail” leads to localized gas channeling, uneven heat transfer, and accelerated erosion on specific tube rows. I’ve seen it; the tube failures show up 18–24 months into operation and the root cause takes a while to trace back.

Protecting Open Connections During Construction

Every open tube end, drum manway, and header hand-hole needs a temporary cover from the moment the component arrives on site until the system is sealed for hydrostatic testing. Construction sites generate weld spatter, concrete dust, and grit continuously. Any of that material entering a steam drum or header will cause pitting corrosion on standing water during pre-commissioning, and tube blockage during initial operation. Use properly fitted steel or heavy rubber blanking caps — not rags, not plastic sheeting taped across the opening. This sounds basic. It gets neglected on busy sites more often than it should.

Pressure Parts Welding, NDT Inspection, and Hydrostatic Testing Protocols

This is the phase where a WHRB project either earns its operational life expectancy or accumulates hidden liabilities that surface six months into service — usually at the worst possible moment. Pressure part integrity is not a quality checkbox. It is the legal and technical foundation on which every subsequent commissioning step depends.

Welder Qualification and Weld Traceability

Before any production welding starts on pressure-retaining components — drum seams, header butt welds, nozzle attachments, tube-to-header connections — every welder must hold a current qualification to the governing code. In practice that means ASME Section IX for projects destined for North America or clients specifying ASME, EN ISO 9606-1 for European and many Southeast Asian installations, or the equivalent national standard (GB/T 15169 in China, for instance). Qualification test pieces must match the actual base material P-number grouping, thickness range, and welding position of the production joints. A welder qualified on thin-wall SA-106B pipe in the 1G position cannot simply be reassigned to a 60 mm drum shell in a 2G position without a separate qualification record.

Traceability is equally non-negotiable. Every weld joint in the pressure circuit needs a unique weld map number, and the weld log entry for that joint should record the welder ID, date completed, filler material heat and batch number, preheat temperature achieved (measured by contact thermometer or temperature-indicating crayon, not estimated), and interpass temperature if the material requires it. For chrome-moly alloy steels — which appear in WHRBs operating above roughly 480°C — post-weld heat treatment records tie into the same log. Lose that paperwork and you will be excavating welds at the third-party inspector’s request, which is expensive and demoralizing.

NDT Coverage Requirements

For longitudinal and circumferential seams on steam drums and large-bore headers, 100% radiographic testing (RT) or ultrasonic testing (UT) is standard and required under most pressure vessel codes. RT is still the default for shop-fabricated drum shells; UT phased array becomes more practical for field-welded thick joints where geometry makes film placement awkward. Either way, the interpretation must be done by a Level II or Level III NDT technician to the applicable standard (ASNT SNT-TC-1A or EN ISO 9712).

Fillet welds — nozzle attachment welds, support lug welds, manway reinforcing pad welds — don’t lend themselves to RT. Magnetic particle testing (MT) or dye penetrant testing (PT) applies here, at a minimum sample rate of 20–25% per code cycle, though a reasonable quality engineer will push toward 100% on nozzle welds in high-pressure circuits. In my experience, skipping thorough PT on small-bore nozzle welds is exactly where stress-corrosion cracking initiates in WHRBs that see thermal cycling from intermittent upstream processes.

Hydrostatic Test Execution

The hydrostatic pressure test is not a formality. Most codes — ASME Section I, EN 12952, and their equivalents — require a test pressure of 1.25× to 1.5× the maximum allowable working pressure (MAWP), held for a minimum of 30 minutes. For a WHRB rated at 4.0 MPa MAWP, that means filling and pressurizing to somewhere between 5.0 and 6.0 MPa with water, then walking every joint, flange face, nozzle, and handhole fitting under pressure.

Two calibrated and independently verified pressure gauges are required by essentially every major code — one at the highest point of the system, one near the test pump — because you need to account for static head differences in tall boiler arrangements. Water temperature matters too. Carbon steel pressure vessels should be tested with water above roughly 7°C to stay well clear of the ductile-to-brittle transition temperature; in winter commissioning scenarios this catches plants off-guard, and heating the fill water in advance is a practical precaution worth scheduling.

Any seepage — even a slow weep at a flange — is a test failure. Visible deformation anywhere in the pressure circuit is a serious finding requiring metallurgical review before proceeding.

Weld Repair Protocol

Rejectable indications found during NDT or hydrostatic testing trigger a defined repair sequence, and there are no shortcuts. The defect location gets marked, confirmed by repeat NDT, and fully excavated — grinding back to sound metal. A qualified welder (same code qualification as original) re-welds the joint, using the same filler material specification. That repair weld then receives full NDT coverage regardless of whether the original joint was a 20% sample zone. If the defect was found during the hydrostatic test, the entire affected circuit must be re-pressurized after repair completion.

Only after clean NDT results and a passed hydrostatic test should the project move to refractory lining and casing installation. Applying refractory over uninspected or repaired-but-untested welds buries the problem and makes any future re-inspection essentially impossible without destructive removal of the casing. That sequence discipline — pressure parts fully cleared before enclosure — is one of the clearest indicators of a competent WHRB installation contractor.

Refractory Lining, Insulation Installation, and Gas-Side Sealing

Get this phase wrong and the boiler will underperform from day one — and you won’t always know it immediately. Casing surface temperatures creeping above spec, unexplained tube corrosion appearing 18 months into operation, heat loss figures that never match the design guarantee. In most cases, the root cause traces back to refractory selection mistakes, a poorly executed bake-out, or gas-side sealing that was rushed before final cladding closed everything in.

Matching Refractory Material to Gas Temperature Zones

A WHRB handling exhaust from a cement kiln or electric arc furnace can see inlet gas temperatures anywhere from 900°C to well above 1,050°C at the entry duct. That rules out ceramic fiber blanket as the primary material in the hottest zone — it simply isn’t dense enough to resist the abrasive, high-velocity gas stream, and on its own it lacks the structural integrity for a hard-wearing exposed surface.

For zones above roughly 800°C, dense castable refractory is the right call. A common specification is 70% Al2O3 content with a bulk density around 2,200 kg/m³, which gives adequate thermal mass and decent resistance to sulfur-bearing gases. Below that threshold — transition ducts, rear passes, lower-temperature sections — ceramic fiber blanket modules rated to 1,260°C or 1,430°C become practical. They’re lighter, easier to install around complex geometry, and they shrink the overall insulation thickness needed. The outermost layer, wherever the external casing panel sits, should target mineral wool or microporous board: the goal there is keeping the casing skin temperature below 50°C in normal ambient conditions, which is both a personnel safety requirement and a contractual thermal loss guarantee in most EPC specs.

Don’t try to simplify this into a single-material solution. A well-designed WHRB will have three or four distinct material zones, each transitioning into the next with carefully detailed edge joints.

Anchor Stud Density and Geometry

Castable refractory on water-cooled membrane walls needs a reliable mechanical key. Stud density typically runs between 25 and 40 studs/m², depending on wall height and expected thermal cycling frequency — a boiler that’s started and stopped daily needs higher density than a baseload unit running continuously for months. V-shaped or dovetail anchor geometry is strongly preferred over straight cylindrical studs. The dovetail profile resists the shear forces generated during thermal expansion cycling; straight studs tend to allow delamination at exactly the wrong moment, when the refractory is hottest and most brittle.

Stud material matters too. Austenitic stainless (commonly 309S or 310S) holds up in high-sulfur environments better than mild steel, which oxidizes and loses grip within a year or two at elevated temperatures.

The Bake-Out Curve: Where Shortcuts Get Expensive

Castable refractory contains chemically bound water that must be driven off in a controlled, staged process. Rush it and you get steam pressure building inside the matrix faster than it can escape — micro-cracking, spalling, sometimes complete panel failure on first firing.

A typical bake-out sequence: ambient cure for at least 24 hours after casting (longer in cold weather, and this is often compressed on tight schedules, which is a mistake), then controlled heating at around 50°C per hour up to 300°C with a four-hour hold, followed by a slower ramp of roughly 25°C per hour up to 600°C with another four-hour hold. The slower second ramp is critical — it addresses chemically bound water that doesn’t release until above 400°C. Do not rely on the upstream process to “naturally” bake out the refractory during first operation. That approach produces uncontrolled ramp rates and nearly always causes cracking.

Gas-Side Sealing: The Detail Most Teams Underestimate

Every penetration through the pressure casing is a potential cold air ingress point: soot blower nozzle sleeves, thermocouple bosses, inspection doors, duct expansion joints, instrument ports. Cold ambient air pulling in through a poorly sealed gap doesn’t just represent heat loss — it creates a localized cold zone on the tube surface immediately downstream, dropping metal temperature below the acid dew point of the flue gas. For sulfur-containing gases, that means sulfuric acid condensation directly on tube OD surfaces. This is one of the more common causes of unexpectedly early tube failure in WHRBs serving glass furnaces and certain chemical process exhausts.

High-temperature compressed ceramic fiber gaskets at flanged connections, backed by stainless steel jacketing on expansion joints, handle most penetrations adequately. Soot blower entry points need sliding seals with sufficient annular clearance to allow thermal movement without binding, but tight enough to prevent significant bypass leakage.

Before the external cladding panels go on — and this is the step that gets skipped most often because the schedule is tight — conduct a gas-tightness test. Pressurize the gas side with low-pressure air, or use a smoke tracer, and walk every seam and penetration. Mark, repair, and retest. Once the outer cladding is installed, finding leaks means cutting panels open.

Speaking of cladding: external panel joints should use overlapping Z-profile edges rather than simple butt joints, sealed with high-temperature silicone rated to at least 300°C. Straight butt joints with standard silicone will open up after the first few thermal cycles and stay open.

Auxiliary Systems Integration: Steam Piping, Feed Water, Blowdown, and Instrumentation Hook-Up

Getting the boiler module set and the refractory cured is the visible part of the job. What actually determines whether the unit runs reliably — or trips on its first week of steam production — is the quality of the auxiliary systems integration. This is where overseas EPC projects most often accumulate problems: wrong pipe class specified at procurement, spring hangers shipped without cold-set load documentation, or a blowdown vessel sized for a different operating pressure. Each of those errors is fixable, but fixing them after hydro test is expensive and sometimes embarrassing when a client’s commissioning team is already on site.

Steam Piping Specification and Thermal Expansion Management

Pipe class selection is not a judgment call — it follows directly from the operating envelope established during front-end engineering. For saturated or lightly superheated steam at or below roughly 400°C and 4.0 MPa, ASME B31.1 seamless carbon steel A106-B is the standard workhorse. Once superheat temperatures climb above 450°C, you are into P91 (9Cr-1Mo-V) territory, and the welding procedure, preheat requirements, and PWHT parameters change substantially. Mixing these materials incorrectly — say, welding A106-B fittings onto a P91 header — creates a dissimilar-metal weld that will eventually fail in creep, usually at an inconvenient hour.

Pipe supports deserve as much engineering attention as the pipe itself. A WHRB steam line will grow anywhere from 40 to 120 mm longitudinally between cold ambient and full operating temperature, depending on line length and steam conditions. Spring hangers must be procured with the correct cold-set load already calculated and locked — not set in the field by whoever happens to be available. If the cold-set load is wrong, the pipe in operating condition either lifts off the support entirely or bears down on it, and either way you risk imposing nozzle loads on the boiler drum outlet that exceed the manufacturer’s allowable limits. Drum nozzle overload is one of those failure modes that shows up as a slow leak two years into operation rather than an obvious commissioning defect.

Feed Water System: Pump Selection and Level Control

The feed water system for a WHRB has a characteristic that conventional fire-tube or coal-fired boiler projects don’t force you to confront as sharply: the evaporation rate fluctuates with the upstream process. A cement kiln going through a clinker change or a glass furnace adjusting pull rate will shift the flue gas temperature — and therefore the WHRB’s steam output — sometimes within minutes. The boiler feed pump must be selected with enough flow control range to follow that variation without cavitating. NPSH margin at minimum flow is frequently under-specified; a centrifugal BFP running at the bottom of its curve in a poorly designed system will cavitate intermittently and destroy the impeller within 18 months.

Three-element drum level control — measuring drum level, steam flow, and feed water flow simultaneously — is the correct approach for any WHRB above roughly 10 t/h capacity. Single-element level control on a variable-load heat recovery boiler is asking for drum level swings and potential carryover into the steam line.

Blowdown Systems and Water Quality Limits

Continuous blowdown (CBD) and intermittent blowdown (IBD) lines need to be routed to a properly sized blowdown vessel before they reach any drain point. The vessel must be designed to flash the incoming blowdown to atmospheric pressure safely — undersized vessels overpressure; oversized ones sometimes fail to maintain the liquid seal. Boiler water TDS limits vary with steam pressure: at 1.0–2.5 MPa, 2,500–3,500 µS/cm conductivity is typically acceptable; at 6.0–9.8 MPa, you are looking at 500–1,500 µS/cm or tighter, depending on the steam purity requirement downstream. These numbers must be written into the water treatment protocol before startup, not derived later from whatever the water treatment contractor happens to suggest.

Instrumentation Layout and Safety Valve Compliance

Drum level indication needs redundancy. A magnetic float gauge gives the operator a direct visual reference; a differential pressure transmitter provides the 4–20 mA signal for the DCS and the three-element control loop. Running either one alone is acceptable for a brief maintenance window, not as a permanent configuration. Thermocouples should be positioned at the flue gas inlet, at least one mid-section location, and the gas outlet — three points minimum — so that fouling or flow maldistribution can be detected from the temperature profile rather than discovered during an unplanned shutdown. An oxygen analyzer at the gas outlet gives you a real-time performance check: if O₂ rises while inlet gas temperature is stable, heat transfer surface fouling is the first thing to investigate.

Safety valves are the one component where field improvisation is genuinely dangerous. Set pressure must comply with the applicable code — ASME Section I for most markets in the Americas, PED/EN ISO 4126 for Europe — and the set pressure stamping must match the boiler design pressure documentation. The cold-set lift test should be conducted and witnessed at the manufacturer’s test bench before the valve ships. On site, if an inspector finds the set pressure has been adjusted without a certified inspector present, it is a code violation in most jurisdictions, and in some markets it voids the pressure vessel registration. It is worth being blunt about this: do not touch the adjusting ring on a safety valve in the field unless a qualified, certified person is standing next to you with the paperwork ready.

Bypass Damper, Soot Blowing, and Online Cleaning System Commissioning

These systems rarely get the attention they deserve during EPC contract negotiations, and that’s where projects get into trouble. A bypass damper that’s undersized or poorly interlocked, or a soot blower sequence that was never properly set up during commissioning, will cost you far more in unplanned downtime than the few days it takes to get them right before first fire.

Bypass Damper Commissioning

The bypass damper is the WHRB’s primary protection device — full stop. When the upstream process trips, surges, or simply goes through a planned startup, you need hot flue gas diverted around the boiler fast and reliably. Actuator travel time matters here: full stroke from open to closed (or reverse) should fall in the 30–60 second range for most industrial WHRBs, though the exact target depends on how quickly your upstream process can generate a thermal transient. A cement kiln bypass gas application is different from a gas turbine HRSG in this regard. Verify actual travel time under load conditions, not just in bench testing — pneumatic actuators in particular can behave differently once they’re fighting against real duct pressure.

Sealing performance at the boiler inlet damper is often under-specified. You want leakage Class IV or better per EN 1751. Anything worse and you’re getting hot gas bleed-through during bypass operation, which means the boiler is seeing heat it shouldn’t be, and your tube metal temperatures will drift upward slowly and invisibly. Request the damper manufacturer’s leakage test certificate and verify it against your actual duct dimensions and operating pressure differential before installation acceptance.

The interlock logic between the bypass damper and the upstream process DCS is where integration errors tend to hide. Walk through every trip scenario — process emergency stop, high gas temperature alarm, boiler drum level low-low, induced draft fan failure — and confirm that the damper response is correct in each case. Don’t rely on the controls vendor’s written logic alone. Physically simulate each trigger and watch what happens.

Soot Blower Setup and Sequencing

For medium-dust applications like cement clinker cooler exhaust or biomass combustion gas, retractable long-range blowers are the standard choice, typically spaced at 1.5–3.0 m intervals across the tube banks. The exact spacing should have come from the thermal-hydraulic design, but it’s worth confirming during commissioning that lance insertion depth actually reaches the design penetration — mechanical stops and guide bearings sometimes shift during transport and erection.

Blowing medium pressure matters more than people think. Steam soot blowers typically operate at 0.8–1.2 MPa; compressed air systems at 0.6–0.8 MPa. Too low and you’re not cleaning effectively. Too high and you risk erosion of tube fins or refractory near the lance entry points, particularly in the first few tube rows where gas temperatures are highest. Check the actual header pressure at the blower inlet under operating conditions, not just at the supply source.

Fouling-Based Blowing Schedules and Ash Handling

Blowing frequency has to be matched to the actual dust loading, not a generic default. High-fouling streams — cement kiln bypass gas, MSW incineration exhaust, certain chemical recovery boilers — may realistically need cycles every 4–8 hours. Integrate the blowing sequence into the boiler control system so that rising differential pressure across the tube banks triggers automatic blowing rather than depending on an operator to notice the trend. A 20–30% increase in gas-side pressure drop over the baseline reading (measured at design gas flow and temperature during commissioning) is a practical trigger point for either an intensified cleaning cycle or an inspection outage, depending on how fast the rise is occurring.

For sticky or corrosive ash — waste-to-energy plants and chemical process recovery boilers are the worst offenders — verify that hopper vibrators and ash conveyor chain cleaners are fully operational before you introduce process gas. This is non-negotiable. Ash bridging in hoppers can develop within a few hours of startup, and once you have a blocked hopper with gas temperatures above 400°C, you’re looking at tube overheating and potential pressure part damage. In practice, I’d run the ash extraction system continuously for at least a shift before first fire just to confirm it’s working cleanly.

Record the baseline differential pressure across each tube bank section at commissioning. That number becomes your long-term fouling reference — without it, trending is guesswork.

Cold Commissioning, First Startup Procedure, and Performance Acceptance Testing

By the time you reach commissioning, months of civil work, welding, and inspection are behind you. This is also where projects most commonly slip — not through catastrophic failure, but through skipped checklist items, undertrained local operators, or pressure to rush startup to meet a commercial deadline. Resist that pressure. A WHRB that enters service before its control logic and water chemistry are properly verified will usually find a way to remind you.

Cold Commissioning: Loop Checks Before Any Heat

Cold commissioning means every system is verified against the P&ID with zero process gas flowing and zero heat applied. Go valve by valve. Every manual isolation, control valve, and safety relief path should be physically walked down and position-confirmed against the marked-up P&ID — not assumed correct because the mechanical contractor signed off. Instrument calibration certificates should be physically on-site and cross-checked against the transmitter tag numbers on the loop drawings. Pressure transmitters on drum level are particularly critical; a miscalibrated drum level instrument is one of the fastest ways to damage an evaporator bank.

Run pump rotation checks with the pump coupling disconnected first, then coupled. Check that all motor phase sequences are correct before full energization. With the DCS or PLC commissioned, inject simulated signals to test every interlock and trip function — high drum pressure trip, low-low water level trip, bypass damper failsafe open on loss of signal — and document the actual response times against the design requirements. This is not optional, and it should not be delegated entirely to the controls contractor without a process engineer in the room watching the logic execute.

Chemical Cleaning and Waterside Passivation

Before the first steam raise, the waterside of every new WHRB needs an alkaline boilout. Mill scale, construction debris, welding flux residue, and whatever grease the erection crew introduced into the drum all need to come out now, not after six months of operation. A typical protocol runs an Na₃PO₄ solution at roughly 0.2% concentration, circulated or boiled for 20–24 hours at low pressure, followed by a full drain, hot water flush, and refill with treated make-up water.

Feed water quality targets for the refill are not suggestions. For units operating above 3.8 MPa, you’re typically targeting dissolved O₂ below 7 ppb, total hardness under 0.03 mEq/L, and SiO₂ below 0.02 mg/L. These numbers shift depending on design pressure — a 1.0 MPa WHRB on a cement kiln has more tolerance than a 9.8 MPa unit behind a gas turbine. Get the water treatment contractor to provide certified lab analysis of the first charge before you proceed.

First Startup: Thermal Ramp and Drum Monitoring

If new refractory was installed, the startup curve is governed by the bake-out requirements of the lining — typically a slow ramp over 48–72 hours with defined hold points. Even where refractory is minimal, thermal shock to headers and drum connections is a real risk. Open the inlet gas damper incrementally, not fully in one motion. The standard working limit most manufacturers specify is a drum metal temperature differential — top shell versus bottom shell — of no more than 50°C during warm-up. Exceed that regularly and you’ll eventually see fatigue cracking at drum nozzle welds.

Monitor flue gas inlet temperature at the WHRB entrance. On processes like EAF steelmaking, the gas temperature is not perfectly steady; there are heat cycles. Your DCS trend logs from this period are worth keeping. They establish a baseline for what normal looks like, which becomes useful the first time an operator calls you at midnight saying something looks wrong.

Performance Acceptance Test

Once the unit has been running steadily at design conditions for a reasonable stabilization period — usually 4–8 hours minimum before data collection begins — you execute the Performance Acceptance Test. Record actual steam output in t/h, drum pressure, superheated steam temperature (if applicable), feed water flow rate, and flue gas outlet temperature at the stack exit. Calculate thermal efficiency from the measured values and compare against the guaranteed figures in the supply contract.

The standard acceptance band in most contracts is ±3% of guaranteed evaporation capacity at 100% design gas flow and temperature. If the measured output falls outside that band, the first things to check are flue gas flow measurement accuracy, feed water temperature entering the economizer, and whether any soot accumulation on the gas side is already affecting heat transfer — possible if the unit has been running on a high-particulate process without soot blowers operating correctly.

Safety valve witness testing should happen during this phase, with the boiler inspector or third-party authority witnessing the lift and reseat pressures physically. Do not accept a paper certification as a substitute.

Commissioning Documentation and Regulatory Handover

The commissioning completion report is a package, not a summary memo. It should contain all DCS trend logs from startup and the PAT, calibration certificates for every pressure, temperature, and flow instrument, NDT reports (radiography or TOFD for pressure-part welds), the hydrostatic test certificate, safety valve witness test records, refractory bake-out logs, and the water chemistry analysis results from the passivation fill.

In overseas EPC projects, this package almost always goes to the client’s insurer and the local boiler authority before commercial operation is permitted. Missing a single document — a common casualty when erection subcontractors are demobilizing quickly — can delay plant startup by weeks while someone hunts down an original calibration certificate or a witness report. Organize this folder as you go, not at the end.

Post-Commissioning Maintenance Schedule and Long-Term Reliability Management

A WHRB that passes performance acceptance testing is not a finished project — it’s the start of a 20- to 30-year operating commitment. In practice, the plants that get the most out of these units are the ones that treat maintenance as an engineering discipline rather than a reaction to breakdowns. The ones that don’t usually find out the hard way, often around year four or five when a tube fails during peak production and the bypass damper becomes the only thing standing between them and a full process shutdown.

Daily and Weekly Operator Rounds

Daily rounds should not be a walkthrough formality. Operators need to log drum level, steam pressure, feed water conductivity, and flue gas outlet temperature — and more importantly, they need to be watching trends, not just point readings. A flue gas outlet temperature that creeps up 8–12°C over two weeks, with no change in upstream process load, is almost always telling you that fouling is building on the tube bundle. Catching it at that stage costs you a planned soot blower cycle. Missing it costs you a tube overtemperature event.

Weekly blowdown and water sample analysis is non-negotiable. Surface blowdown controls dissolved solids near the waterline; bottom blowdown removes settled sludge from the drum. Both need to be done on a fixed schedule, not when someone remembers.

Monthly Mechanical Checks

Soot blower lance condition deserves a hands-on inspection every month in high-dust gas streams — cement kiln exhaust, EAF off-gas, and similar applications. Lances that have developed pitting or scale buildup from the steam itself deliver uneven cleaning coverage, which means some tube rows stay fouled while others get blasted clean. Over time that creates thermal asymmetry across the bundle. Check actuator lubrication on bypass dampers at the same interval; a damper that sticks open or closed under emergency conditions is a serious operational risk.

Annual Internal Inspection and Tube Thickness Monitoring

The annual internal inspection is where you actually learn what your boiler has been doing. Pull the inspection covers, get someone inside the drum (with proper confined space procedures), and look at the tube surfaces. Erosion, corrosion pitting, and fouling deposits all tell a story about operating conditions that your instrumentation can only approximate.

Ultrasonic thickness gauging at predetermined grid points is the only reliable way to track erosion quantitatively. In high-dust gas streams, erosion rates of roughly 0.1–0.3 mm/year are common — the actual rate depends on gas velocity, particle loading, and tube material. Map those measurements against your original wall thickness specifications and trend them year over year. That trending is what lets you predict remaining service life instead of guessing.

Water chemistry management is where a lot of plants cut corners and eventually pay for it. Scale deposition raises tube metal temperatures into creep risk territory in high-pressure superheater sections faster than most operators expect. Maintain dosing discipline, keep blowdown records, and check with your water treatment supplier if conductivity trends start moving in the wrong direction seasonally — raw water quality in many regions shifts significantly between dry and wet seasons, and your dosing program needs to account for that.

Statutory Outage Planning and Spare Parts Inventory

Most jurisdictions require a formal pressure vessel inspection by an authorized inspector every 2–4 years. Plan for this outage, don’t just react to the certification deadline. Use the window to replace soot blower packing sets, repair any refractory damage found during internal inspection, swap out thermocouples that have drifted (they drift more than most people expect after 18–24 months at high flue gas temperatures), and clean tube surfaces properly.

For plants running WHRBs sourced from manufacturers in China, lead times for replacement components — tube sections, gaskets, safety valve trim kits, bypass damper actuator seals — can realistically run 8–16 weeks depending on the component, port congestion, and customs clearance. Keep a working spare parts inventory on site: at minimum, tube samples of each specification in the unit, a full set of manway and hand-hole gaskets, and actuator seal kits. The cost of holding that inventory is trivial compared to the production loss from a three-month wait on a single fitting.

Frequently Asked Questions About Waste Heat Recovery Boiler Installation

How long does a WHRB installation actually take?

The honest answer depends almost entirely on unit size and local permitting pace — not on how quickly the erection crew works. A skid-mounted, factory-assembled unit in the 2–5 t/h range can realistically be mechanically complete in 8–10 weeks from site mobilization. A field-erected unit in the 50–150 t/h class typically runs 20–24 weeks, sometimes longer if the local boiler inspection authority has a backlog for witness testing.

The critical path is almost never the steel erection or pipe spooling. It’s usually the civil foundation cure time (typically 28 days minimum for reinforced concrete to reach design strength), the NDT queue after pressure-parts welding, and — in many overseas markets — getting the statutory inspection body on-site for the hydrostatic test witness. Procurement teams that underestimate the inspection scheduling window routinely cause two-to-four-week delays on otherwise well-run projects. Build that buffer in from week one.

Can the upstream process keep running during installation?

Most of it, yes. The structural steel, pressure-part assembly, refractory work, and auxiliary hook-ups can all proceed while the kiln, furnace, or turbine keeps running. The unavoidable process tie-in is at the flue gas duct — both the inlet connection and the outlet or bypass branch. That physical cut-in almost always requires a planned shutdown, typically 24–72 hours depending on duct size and access.

Prefabricating duct spools to near-final dimensions off-site and pre-fitting a bypass damper frame in parallel with the main installation are the two most effective ways to compress that shutdown window. In practice, a well-organized site team can execute the duct tie-in during a scheduled maintenance stop rather than calling a dedicated outage. That requires upstream coordination starting months before the hot-work date, not weeks.

What installation errors cause the most early failures in service?

Five mistakes account for the majority of infant-mortality failures seen in WHRBs during the first two years of operation:

1. Wrong sliding/fixed bearing arrangement. Locking a header or drum that should float axially forces constrained thermal expansion into the tube rows and weld joints. Header cracking typically shows up within the first few thousand operating hours.
2. Poor gas-side casing sealing. Air ingress through poorly sealed expansion joints or penetrations drives acid dew-point corrosion on economizer tubes, especially with moderate-to-high sulfur fuels. The damage is slow and invisible until you have a tube leak.
3. Skipping or rushing the refractory bake-out curve. Castable refractory contains significant moisture. Fire it too fast and the steam pressure fractures the lining, exposing pressure-part surfaces to direct flame impingement or corrosive gas channeling.
4. Inadequate water treatment at initial startup. New boilers accumulate mill scale, welding flux residue, and preservation compounds during fabrication. Without proper chemical cleaning and controlled first-fill procedures, scaling starts immediately on the hottest tube surfaces.
5. Undersized or incorrectly set safety valves. This is a regulatory compliance failure as much as a mechanical one. It shows up in inspection audits and can result in forced shutdown.

Which codes and certifications govern WHRB installation in different markets?

Taishan Group’s WHRBs carry ASME, CE, and Chinese GB certifications. For specific project jurisdictions, documentation packages and third-party inspection coordination can be adapted accordingly.

At what flue gas temperature does a WHRB become economically viable?

The rule of thumb most experienced engineers use: above roughly 300°C inlet temperature, a steam-generating WHRB pays back at acceptable rates under most industrial energy tariff structures. Below 300°C, the heat duty shrinks fast, and the capital cost of pressure parts becomes hard to justify — an economizer-only configuration for feed water preheating, or an ORC system for power generation, usually makes more sense.

One number that matters more than people expect is the acid dew-point temperature of the specific flue gas. For low-sulfur natural gas exhaust it sits around 50–60°C, which is fairly benign. For high-sulfur fuels — certain petroleum cokes, high-S coal — it can reach 130–160°C, which sets a hard floor on how cold you can cool the exit gas. That floor directly determines how much heat is actually recoverable, and ignoring it in the initial sizing calculation is a mistake that no amount of clean execution can fix later.

Does Taishan Group offer full EPC delivery for WHRB projects?

Yes. The full scope covers process thermal design, WHRB engineering and manufacturing, auxiliary system supply (feed pumps, deaerators, steam headers, blowdown vessels, instrumentation, and control systems), site installation supervision, commissioning, operator training, and ongoing after-sales technical support. For overseas industrial and power projects, Taishan Group has experience managing the documentation, inspection, and logistics coordination that EPC delivery in export markets actually requires — not just the boiler itself.

References

1. National Board Inspection Code — Part 1: Installation
   Source: The National Board of Boiler and Pressure Vessel Inspectors

2. ASME Boiler and Pressure Vessel Code — Boiler Requirements and Guidance
   Source: American Society of Mechanical Engineers (ASME)

3. National Board Jurisdictional Boiler and Waste Heat Boiler Requirements
   Source: The National Board of Boiler and Pressure Vessel Inspectors

4. National Board Boiler Owner and Operator’s Guide
   Source: The National Board of Boiler and Pressure Vessel Inspectors

5. NBIC Part 1 Installation Training for Boilers and Pressure Equipment
   Source: The National Board of Boiler and Pressure Vessel Inspectors

6. 2025 NBIC Installation Requirements and Major Changes
   Source: The National Board of Boiler and Pressure Vessel Inspectors

7. Commissioning Process for Federal Facilities
   Source: U.S. Department of Energy (DOE)

8. Industrial, Commercial, and Institutional Boilers and Process Heaters: NESHAP for Major Sources
   Source: U.S. Environmental Protection Agency (EPA)

9. Compliance for Industrial, Commercial, and Institutional Area Source Boilers
   Source: U.S. Environmental Protection Agency (EPA)

10. NESHAP for Area Source Boilers Tune-up Guides
    Source: U.S. Environmental Protection Agency (EPA)