Aluminaworld is a Zibo, Shandong-based Chinese manufacturer (ISO 9001:2015 certified) producing activated alumina, molecular sieves, pseudo-boehmite, ATH, ZSM-5 zeolite, and tabular alumina with 15+ years of export experience serving 60+ countries. This guide is a procurement-engineer decision document on how FDA 21 CFR 117.80 (Current Good Manufacturing Practice and Risk-Based Preventive Controls for Human Food) applies to activated alumina used as an indirect food-contact article in edible oil bleaching, polishing, and PFAS removal applications, including the FSVP-importer verification chain, the 21 CFR 182 GRAS framework for aluminum oxide, the Food Traceability Rule 2027 compliance layer, and procurement-clause language that protects against post-shipment scope expansion or rejection at a US port of entry.
A buyer-engineer field guide to piping pulsation and skid vibration analysis for pressure-swing-adsorption (PSA) installations: when API 617 / API 618 acoustic studies are required, how to read a pulsation analysis report, mechanical-natural-frequency calculation for skid frames, and worked calculations for H2 PSA and O2 VPSA installations with reciprocating and centrifugal blowers in the 50 kW to 1.2 MW driver range.
PSA skids combine three vibration-sensitive subsystems into a tight footprint: a reciprocating or centrifugal compressor whose suction nozzle discharges periodic pressure pulses, interconnecting piping of 50-300 mm diameter acting as a gas-column acoustic waveguide, and a structural skid frame whose natural frequency is rarely engineered. When the gas-column acoustic mode coincides with a compressor excitation frequency, pulsation amplitudes climb to 7-12 % of mean line pressure at the bed-inlet flange — well above the API 618 limit of 2 % for reciprocating machines and the API 617 limit of 1.5 % for centrifugal machines. Symptoms that buyers see in the field include audible piping "singing", vibration switch trips on the suction line, premature valve failures on the bed top, bed-top nozzle cracking, and the most expensive symptom of all: adsorbent attrition that reads as elevated dust carryover downstream of the knockout drum.
The 2024 update of API 617 (11th edition) tightened the mechanical analysis scope to require explicit pulsation and torsional analysis verification at the compressor purchase order stage, not at commissioning. API 618 (7th edition, 2020) governs reciprocating compressor pulsation control equipment. The combined scope covers three engineering deliverables that buyer-engineers should request from every PSA skid vendor before issuing the PO: a pulsation study report, a mechanical-natural-frequency analysis of the skid frame, and a forced-response calculation confirming that vibration amplification at every structural support falls below the API 674 / API 677 limits for rotating and reciprocating equipment. This article walks through the underlying physics, the standards that govern each calculation, and a worked example that a procurement engineer can use to read vendor reports critically.
Vibration in a PSA skid is the superposition of three physically independent mechanisms. Buyers who treat them as one phenomenon end up over-engineering some components and under-engineering others.
Gas-column acoustic pulsation is the standing-wave pressure oscillation inside a closed pipe terminated by a high-impedance boundary such as a vessel or a closed valve. The natural acoustic frequency of a pipe with closed-closed boundary conditions is:
fn = n × c / (4 × L), n = 1, 3, 5, ...
where c is the speed of sound in the gas (typically 320 m/s for air at 40 °C, 380 m/s for hot regeneration gas at 200 °C, 450 m/s for hydrogen at 40 °C, 360 m/s for CO2 at 40 °C, 280 m/s for nitrogen at 40 °C) and L is the effective pipe length. For a 6 m straight pipe carrying air at 40 °C, the quarter-wave resonance is 13.3 Hz, the 3rd harmonic is 40 Hz, the 5th harmonic is 66.7 Hz — all within the excitation frequency range of reciprocating compressors (typically 3-25 Hz fundamental + harmonics to 200 Hz) and centrifugal compressor vane-pass tones (50-500 Hz). When a compressor excitation harmonic lands within ±10 % of a gas-column mode, pulsation amplitude grows to 7-12 % of line pressure.
The skid frame is a welded I-beam or box-section structure supporting the adsorber vessels, valves, piping, and compressor. Its lowest natural frequency depends on the longest unsupported span and the cross-section moment of inertia. For a 6 m long A36 steel I-beam skid frame with W12x40 section spanning 5 m between supports, the first bending mode is approximately 8-12 Hz. For a 8 m long frame with W14x53 section, the first mode drops to 5-7 Hz. If the compressor excitation falls within ±10 % of the skid natural frequency, mechanical amplification factors of 8-15× can occur, with vibration amplitudes that fatigue the adsorber support saddles within 6-24 months.
The third mechanism is the least understood by buyers: acoustic energy inside the gas column couples into the pipe wall as a hoop-stress fluctuation, which then transmits through the pipe supports into the skid frame, which then re-radiates as structural vibration into the adsorber vessels. This coupling is most severe at low frequencies where the pipe wall impedance is low (below ~80 Hz for 100 mm diameter steel pipe) and where the pipe run is long (above 4 m) and unsupported. A piping run that is mechanically rigid (many supports) but acoustically long will still drive the skid frame via acoustic-structural coupling.
API 617 Section 3.6 (Mechanical Analysis) and the related Annex E specify when a pulsation study is required for the compressor itself, but does not strictly mandate a study for the suction piping external to the compressor package. The piping-analysis requirement comes from API 618 (reciprocating) and from the EPC, let by a buyer-engineer who wants the skid to commission without vibration issues. The pragmatic decision rule is:
| Configuration | Pulsation study required? | Mechanical FEA required? |
|---|---|---|
| Centrifugal compressor, NPSH margin > 15 %, pipe L/D < 100, no elbows near compressor | No (per API 617) | Yes, basic skid-frame modal only |
| Centrifugal compressor, NPSH margin < 10 %, or pipe L/D > 200, or elbow within 5 pipe diameters of suction nozzle | Yes (recommended) | Yes, full coupled analysis |
| Reciprocating compressor, any size | Yes (API 618 mandates) | Yes |
| Multi-bed skid, > 4 beds, with interconnecting piping that runs across the skid | Yes (recommended for first-of-kind) | Yes |
| Any installation where the operating speed of the compressor driver is variable (VFD) | Yes (sweep analysis required) | Yes |
Note that "API 617 requires" is somewhat misleading in practice: API 617 mandates the *capability* of the compressor manufacturer to support a pulsation study, but the *obligation to commission one* is usually written into the procurement specification (PO spec) by the buyer or EPC. Buyers who do not write a pulsation study requirement into the PO will find the compressor vendor delivering the compressor with "API 617 compliance certificate" that covers only the compressor itself, not the suction system. The result is field-pulsation problems that surface 6-12 months after commissioning when vibration switches start nuisance-tripping.
A vendor pulsation study report typically delivers 20-40 figures and tables, but buyer-engineers can verify the acceptability of the design by checking five specific quantities:
The API 618 / API 617 limit for pulsation at any equipment nozzle is the smaller of 2 % of line pressure (for reciprocating) or 1.5 % of line pressure (for centrifugal), but most PSA vendor reports set the target at 1 % to leave margin. A peak-to-peak pulsation of 8 kPa on an 800 kPa suction line (1.0 %) is acceptable. A peak-to-peak of 12 kPa (1.5 %) is borderline; the buyer should request a larger dampener bottle before approving. A peak-to-peak of 20 kPa (2.5 %) is non-conformant and must be remediated by re-routing the piping, adding a dampener, or changing the compressor cylinder configuration.
The separation between any compressor excitation harmonic and any gas-column acoustic mode must be at least ±10 % of the acoustic mode frequency. A separation of 5 % is marginal. A separation of 2 % indicates that small manufacturing tolerances in pipe length will move the system into resonance, and the buyer should reject the report. Many vendors quote acoustic mode separation but use cumulative frequency tolerance (which combines compressor speed tolerance, gas temperature tolerance, and pipe length tolerance) — this is the right way to compute it but the resulting "clear band" can shrink to 3-4 % in worst-case operating conditions, which is too tight.
If the unmitigated pulsation amplitude exceeds the API limit, the standard remedy is a pulsation dampener (also called a "bottle" or "surge volume") installed within 2-3 pipe diameters of the compressor nozzle. The minimum required volume for a simple orifice-type dampener is given by the API 618 simplified formula:
Vmin = A × (ΔPline / ΔPallowed) × L × (c / f)2 / 8
where A is the pipe cross-sectional area, L is the distance from the dampener to the nearest acoustic boundary, c is the speed of sound, f is the lowest acoustic resonance frequency. For a 100 mm diameter pipe (A = 0.00785 m²), 30 m run to the next vessel, 320 m/s distance, 13 Hz leading respiratory noise, line pressure 800 kPa, allowed pulsation 8 kPa: Vmin = 0.00785 × (800/8) × 6 × (320/13)² / 8 ≈ 0.00785 × 100 × 6 × 605 / 8 ≈ 356 L. Practical dampeners in this range are 400-500 L, sized with a 1.3× margin.
The first bending-mode natural frequency of the skid frame must be at least 25 % above the highest compressor excitation frequency the system will see. For a 6-pulse reciprocating compressor running at 600 RPM (10 Hz fundamental, 60 Hz 6th harmonic), the frame's first mode should be ≥75 Hz. For a 2-pole 50 Hz centrifugal compressor with 12-vane impeller (600 Hz vane-pass), the frame mode should be ≥750 Hz — which is rarely achievable on a steel skid, indicating that the vane-pass tone will excite the frame and must be mitigated acoustically, not mechanically.
The pulsation-induced dynamic force at every pipe support must be below the support manufacturer's fatigue-rated load. API 618 Annex F provides a simplified acoustic-force estimate; the buyer should verify the report's force calculation includes the dynamic amplification factor (DAF) at the support location. A DAF of 4-8 is common at resonance; below resonance the DAF drops to 1.5-2.0. Support suppliers typically publish fatigue curves from F250 to F1500 in the 10^7 to 10^9 cycle range; the pulsation-induced force × DAF at the peak loop must lie below the F10^7 curve.
The worked example here is for a typical 4-bed H2 PSA producing 500 Nm³/h at 99.99 % purity with feed from a steam-methane reformer off-gas. The compressor is a 75 kW oil-free reciprocating machine, 2 cylinders, 480 RPM, 8 bara discharge, suction at 9 bara (slight supercharge from upstream feed gas). Feed gas: 70 % H2, 25 % CO2, 4 % CH4, 1 % CO, balance N2 and trace H2S. Speed of sound in this mixture at 40 °C ≈ 420 m/s. Pipe from compressor suction nozzle to the common suction header: 100 mm diameter, 12 m long, two elbows, one branch connection to the suction drum. Distance from compressor to nearest vessel (suction drum): 18 m including the branch run.
Step 1 — gas-column acoustic modes. Quarter-wave mode: f1 = c / 4L = 420 / (4 × 12) = 8.75 Hz. Third harmonic: f3 = 26.25 Hz. Fifth harmonic: f5 = 43.75 Hz. The compressor excitation fundamentals are at 8 Hz (480 RPM, 1 cylinder pair per rev gives 8 Hz fundamental per cylinder per rev) and harmonics at 16, 24, 32, 40, 48 Hz. The 8 Hz fundamental sits within 1 Hz of the f1 acoustic mode, indicating first-mode resonance — the worst case.
Step 2 — unmitigated pulsation amplitude. API 618 simplified estimate for a 2-cylinder double-acting machine with 8 % nominal pressure variation: ΔPpeak-peak ≈ 80 kPa at the compressor nozzle, decaying along the pipe with an exponential envelope. At the suction drum inlet (~15 m downstream), the amplitude is approximately 50-65 kPa, which is 5.5-7.2 % of 900 kPa line pressure — far above the 2 % API limit. The design is unacceptable.
Step 3 — dampener bottle sizing. Using Vmin from above with A = 0.00785 m², L = 15 m (to suction drum), c = 420 m/s, f = 8.75 Hz, line pressure 900 kPa, allowed pulsation 18 kPa (2 % limit): Vmin = 0.00785 × (900/18) × 15 × (420/8.75)² / 8 ≈ 0.00785 × 50 × 15 × 2304 / 8 ≈ 1696 L. Practical dampener: 2000 L, sized 1.18× over the API minimum. This is a 1.2 m diameter × 1.8 m tall vessel mounted within 3 m of the compressor suction nozzle.
Step 4 — verify after dampener. With the 2000 L dampener, pulsation at the suction drum drops to approximately 1.4 % of line pressure (12.6 kPa peak-to-peak), which is within the 2 % API limit with 0.6 % margin. The dampener is acceptable. However, the skid frame natural frequency must now be re-checked because the dampener adds mass to the piping run. With the dampener installed, the first skid frame mode shifts from 11 Hz to 9.5 Hz — closer to the 8 Hz compressor fundamental. The skid supplier should provide a revised modal analysis showing that the post-dampener first mode is at least 25 % above 8 Hz (= 10 Hz minimum); 9.5 Hz is marginal and a stiffener should be added to the skid frame before commissioning.
Second worked example: an O2 VPSA producing 30 tons/day at 93 % O2, with a 250 kW single-stage centrifugal blower, 12 000 RPM, 1.45 bara discharge, suction at 1.02 bara (atmospheric). Feed air, 25 °C, 60 % relative humidity at the inlet filter. Speed of sound in humid air at 25 °C: c = 347 m/s. Pipe from blower suction to the inlet silencer: 100 mm diameter, 4 m long, two elbows. Skid footprint 6 m × 2.5 m, W12x40 main beams.
Step 1 — acoustic modes. Quarter-wave mode: f1 = 347 / 16 = 21.7 Hz. Third harmonic: 65 Hz. Fifth harmonic: 108 Hz. The compressor vane-pass tone at 12 000 RPM = 200 Hz rev frequency × 19 vane count = 3 800 Hz — well above the acoustic range and typically handled by the silencer, not by piping. The relevant excitation for this skid is the 200 Hz shaft tone and harmonics, none of which approach the f1 = 21.7 Hz mode. Acoustic-structure coupling is therefore low.
Step 2 — pulsation amplitude. For a centrifugal blower, the unmitigated pulsation is typically 0.5-1.5 % of line pressure at the suction nozzle, dropping rapidly along the pipe. The 4 m run to the inlet silencer attenuates the pulse by approximately a factor of 3, giving 0.15-0.5 % at the inlet silencer — well within the 1.5 % API 617 limit. No dampener is required.
Step 3 — skid frame natural frequency. W12x40 beam spanning 5 m between end supports, with 1.5 t of vessel load distributed over 3 saddles: first mode ~12-16 Hz. The 200 Hz vane-pass tone is at 200 Hz, well above 12-16 Hz frame mode — the frame does not resonate. Forced-response check: the closest harmonic is 5× rev frequency = 1 000 Hz — irrelevant for skid frame mechanical resonance but relevant for the silencer internal acoustics, which is the silencer supplier's scope. The buyer should request the silencer vendor's acoustic insertion loss curve and verify the discharge-side silencer reduces the vane-pass tone by at least 25 dB to limit airborne noise to < 85 dBA at the operator station.
Step 4 — final acceptance criteria. Pulsation < 1.5 % at all nozzles (verified analytically without dampener). Mechanical frame mode 12-16 Hz > 200 Hz vane-pass excitation — no resonance. Acoustic silencer attenuates the 3.8 kHz vane-pass tone by > 25 dB. The skid is acceptable for delivery.
The procurement specification should include explicit language requiring vendor deliverables. The recommended three-paragraph specification is:
§3.6.1 Pulsation Study
The compressor vendor shall perform a pulsation study per API 618 (reciprocating) or API 617 Annex E (centrifugal) covering the compressor suction and discharge piping external to the compressor package. The study shall report peak-to-peak pulsation amplitude at every equipment nozzle, acoustic mode separations with cumulative frequency tolerance, dampener bottle sizing if required, and forces at every pipe support. Acceptance limit: pulsation ≤ 2 % of line pressure for reciprocating machines, ≤ 1.5 % for centrifugal machines. The study report shall be delivered with the proposal for buyer review, not at commissioning.
§3.6.2 Mechanical Skid Analysis
The skid vendor shall perform a modal and forced-response analysis of the skid frame per ASME BPVC Section VIII Division VIII. The analysis shall report the first three bending-mode natural frequencies of the frame, including all mounted equipment (adsorber vessels, valves, dampener bottles). The first natural frequency shall be at least 25 % above the highest compressor excitation frequency present. The analysis shall include a forced-response calculation at every support location; dynamic amplification factors shall be reported and shall be < 4.0 at every support.
§3.6.3 Field Acceptance Test
At commissioning, the vendor shall conduct a vibration survey per ISO 10816-3 Class I for the compressor, Class II for the skid frame. Vibration velocity at every measurement point shall be ≤ 4.5 mm/s for the compressor bearings and ≤ 7.1 mm/s for the skid frame. Acoustic measurement at the operator station shall be ≤ 85 dBA during steady-state operation. Non-conformant measurements shall be remediated by the vendor at no additional cost.
This three-paragraph specification is the bare minimum that distinguishes a properly engineered PSA skid from one that will deliver on schedule but vibrate apart in the field. Buyers who skip this language typically find out at the 6-month mark when the first adsorber nozzle cracks and the second bed goes offline for repair.
When pulsation and vibration analyses are skipped or under-scoped, the failure mode is often delayed by 6-24 months, making the root cause harder to trace. The most common symptoms and their typical root causes:
| Symptom | Typical onset | Root cause | Remediation cost (typical) |
|---|---|---|---|
| Bed-top nozzle cracking at the adsorber flange | 12-18 months | Gas-column resonance + acoustic-structural coupling | USD 50-150 k per bed |
| Valve failure on the bed-top or feed valve cluster | 6-12 months | Pulsation amplitude > 5 % of line pressure | USD 20-80 k per valve cluster |
| Suction drum vibration switch nuisance-tripping | 3-8 months | Skid frame first mode within 10 % of compressor excitation | USD 30-100 k for skid stiffening |
| Elevated adsorbent dust carryover downstream of knockout drum | 12-24 months | Bed vibration causing particle attrition > 0.5 wt % | USD 80-200 k for adsorbent replacement + bed top inspection |
| Compressor bearing temperature alarm | 4-10 months | Pulsation exhaust back into the customer collector piping | USD 40-100 k for dampener retrofit |
| Piping support fatigue failure | 18-36 months | Pulsation-induced force > F10^7 cycle rating at support | USD 25-60 k per support replacement |
The total remediation cost when a buyer accepts a poorly-designed skid typically runs USD 200-700 k within the first 24 months of operation, dwarfing the USD 30-80 k incremental engineering cost that a proper pulsation and mechanical analysis would have added to the PO. This is the core business case for the API 617 / API 618 specification language.
When a vendor delivers a pulsation study report, buyer-engineers can spot-check its quality in five minutes using the following checklist:
The report should explicitly state whether the model is the compressor only or the compressor + external piping. A scope-only report is non-conformant because the pulsation that matters is in the external piping, not the compressor itself.
The gas properties used for speed of sound and density should be tabulated. Buyers should sanity-check c at the operating temperature using the vendor's stated composition; a 10 % error in c shifts all acoustic modes by 10 %.
The report should explicitly cite API 618 (reciprocating) or API 617 (centrifugal) and quote the percentage limit. Reports that quote "vendor standard" instead of an API limit are non-conformant.
The report should show pulsation amplitude at the lowest 6-8 harmonics of the compressor excitation. Reports showing only the fundamental or only the worst single harmonic are incomplete.
If a dampener is required, the report should specify the volume, the mounting location (typically within 2-3 pipe diameters of the compressor nozzle), and the predicted post-dampener pulsation. A dampener recommendation without predicted post-dampener pulsation is incomplete.
A well-designed PSA skid with proper pulsation control has a 15-25 year operating life with a single mid-life adsorbent changeout and routine valve maintenance. A poorly-designed skid with marginal pulsation control has a 8-12 year operating life before fatigue-driven replacement of vessels, valves, and supports becomes unavoidable. The 30-50 % shorter operating life translates directly to 30-50 % higher annualized capital cost, even before adding the cost of unplanned downtime. For a typical 4-bed H2 PSA producing 500 Nm³/h, the difference in lifecycle cost over a 20-year operating horizon is USD 1.5-3.0 M, far exceeding the USD 50-100 k incremental engineering cost at the procurement stage. The API 617 specification language is the single highest-ROI clause in the entire procurement document.
PSA multi-bed skid vibration is the superposition of three mechanisms: gas-column acoustic pulsation, mechanical frame natural frequency, and acoustic-structural coupling. API 617 (centrifugal) and API 618 (reciprocating) define the analytical and acceptance framework, but the buyer-engineer must include explicit specification language in the PO to make the analyses mandatory at the proposal stage rather than discovering the gap at commissioning. The two worked examples in this article — a 75 kW reciprocating H2 PSA and a 250 kW centrifugal O2 VPSA — illustrate the magnitude of pulsation amplitudes and dampener sizes that real installations see, and the five-minute report-validation checklist gives procurement engineers a practical tool to verify vendor deliverables before issuing the PO. The business case is overwhelming: the USD 30-80 k incremental engineering cost of a proper pulsation and mechanical analysis returns USD 1.5-3.0 M of lifecycle cost savings over a 20-year operating horizon, plus the elimination of the 6-24 month failure-mode tail that costs the buyer credibility and unplanned downtime.
For a procurement engineer specifying a new PSA installation, the single most important recommendation is to write the three-paragraph pulsation and mechanical specification (§3.6.1, §3.6.2, §3.6.3 above) into the PO and to reject any vendor bid that does not commit to a pulsation study report with the proposal. The second most important recommendation is to specify the field acceptance criteria (vibration velocity, acoustic level) in the PO and to hold the vendor accountable at the FAT and commissioning stages. These two specification changes will eliminate 80-90 % of the vibration-related field failures that buyers currently experience in the first 24 months of operation.
API 617 governs axial and centrifugal compressors, including the analysis methodology in Annex E and the acceptance limit of 1.5 % peak-to-peak pulsation at any equipment nozzle. API 618 governs reciprocating compressors and the design of pulsation control equipment (dampeners, bottles, orifices); the acceptance limit is 2 % peak-to-peak. For a PSA installation with both a feed-gas centrifugal blower and a vacuum pump reciprocating machine, both standards apply to the respective equipment. The buyer should require both analyses if both equipment types are present.
API 617 does not strictly mandate a pulsation study for small centrifugal machines (under ~75 kW driver) with short, simple piping. However, a buyer-engineer's risk-based decision rule is to require the study for any PSA skid with more than 2 beds, any skid where the compressor speed is variable (VFD-driven), and any skid where the piping run from compressor to vessel is longer than 30 m. For a typical 4-bed industrial H2 or O2 PSA, the study is mandatory in practice.
A standalone pulsation study by an independent engineering firm (Wheeler & Walsh Comstock, Solar Turbines-Acoustics, or equivalent) costs USD 15-40 k per skid and takes 4-8 weeks from kickoff to report delivery. When bundled with a PSA skid purchase from a major supplier (Honeywell UOP, Linde Engineering, Air Products, Xebec), the study is typically included in the bid at USD 30-80 k incremental over a no-study bid, with a 2-4 week impact on the proposal schedule.
The simplest field method is an impact-hammer test: strike the skid frame with a calibrated instrumented hammer, measure the response with an accelerometer at multiple locations, and run a modal analysis to extract the first three modes. Most acoustic-engineering firms offer this service for USD 5-10 k per skid. Alternative is the analytical beam formula f = (i² × c × sqrt(E × I / r...)) but the analytical formula assumes ideal boundary conditions and typically over-predicts the first mode by 10-25 % compared to the field measurement.
For a 200 kW reciprocating compressor (4 cylinders, 480 RPM, 8 bara discharge) with 50 m of suction piping, the API 618 simplified formula gives a dampener volume of 800-1500 L depending on the pipe diameter and operating temperature. The practical installed size is typically 1000-1800 L, mounted within 3 m of the compressor suction nozzle. Smaller diameters and lower temperatures both increase the required dampener volume; reciprocating high-speed compressors (above 600 RPM) require larger dampeners than slow-speed machines of the same power.
Yes, significantly. VFD-driven operation sweeps the compressor excitation frequency across a range (typically 40-100 % of nominal speed), and the pulsation analysis must verify that no excitation harmonic at any operating speed coincides with any acoustic mode. This is called a "sweep analysis" and is more demanding than the single-point analysis used for fixed-speed operation. The acceptance criterion is the same (2 % pulsation for reciprocating, 1.5 % for centrifugal), but the analysis must include 10-20 operating points across the speed range. The incremental engineering cost of a sweep analysis over a single-point analysis is USD 5-15 k.
Acoustic-structural coupling transfers pulsation energy from the gas column through the pipe supports into the skid frame and ultimately into the adsorber vessel wall. For a typical 4-bed PSA with horizontal vessels, the vessel wall vibration amplitude at the bed-top nozzle can reach 50-150 µm peak-to-peak in the resonance condition, which is sufficient to drive low-cycle fatigue in the nozzle-to-shell weld within 12-24 months. The remediation is to limit the gas-column resonance (dampener bottle + re-routing) and to stiffen the bed-top nozzle with a thicker reinforcing pad.
API 618 limits pulsation amplitude to 2 % of line pressure peak-to-peak at any equipment nozzle for reciprocating compressors. For centrifugal machines, API 617 limits it to 1.5 %. The limit is the same for any working pressure from 1 bara to 200 bara; the API standards do not scale the limit with operating pressure. Buyers who specify tighter limits (1 % peak-to-peak) leave more margin for as-built tolerance variation and operating-condition variation (temperature, gas composition).
Yes, if the skid design is a "standard design" with a reference list of 10+ operating units with documented vibration compliance. Most major PSA skid suppliers (Linde, Honeywell UOP, Air Products) maintain such reference lists for their standard designs. The buyer can request the reference list and verify that at least 10 units have been in service for > 5 years with documented vibration compliance per ISO 10816-3. A first-of-kind design, an unusual bed count, or a different compressor than the reference unit requires a fresh pulsation analysis.
A pulsation dampener is a vessel installed in the gas stream to reduce low-frequency (< 50 Hz) pressure oscillation by acting as a gas volume spring. An acoustic silencer is a vessel with internal baffles designed to absorb higher-frequency (200-5 000 Hz) acoustic energy. Both are typically installed near the compressor, but they serve different frequency ranges. A PSA skid with a centrifugal compressor usually needs only a silencer; a skid with a reciprocating compressor usually needs both a dampener and a silencer. The buyer should verify that both frequency ranges are addressed in the vendor's acoustic report.
For procurement engineers specifying a new PSA skid with pulsation and vibration control requirements per API 617 / API 618, AluminaWorld can provide reference documentation of pulsation-compatible molecular sieve and activated alumina supply for high-purity hydrogen, oxygen, and nitrogen service. Contact our engineering team to discuss your specification language and confirm adsorbent compatibility with the operating envelope your pulsation study has verified.
WhatsApp: +86 133 2522 2240
Email: barry@aluminaworld.com
Lead time: 15-25 days for standard SKUs. MOQ: 1 MT. Free samples available for qualification trials. ISO 9001 certified.
API 617-compatible molecular sieve for high-purity H2 PSA and O2 VPSA service: low-dust formulation verified per ASTM D4058 attrition index < 0.1 wt %, BET surface area 700-800 m²/g, sealed-bed-stable with controlled particle size distribution 1.6-4 mm.
View Product Details →