Compressor Foundation Grouting and Alignment Interfaces

Compressor Foundation Grouting and Alignment Interfaces

Not an approved installation procedure or evidence of supplier approval. All numerical worked examples are illustrative, not universal tolerances.

1. Direct engineering answer and controlled ownership

Compressor foundation integrity cannot be established by a level reading or a grout pour alone. The release decision requires an unbroken evidence chain from machine configuration and OEM support reactions through civil design, prepared concrete, anchor and grout installation, machine support condition, coupling or driven-equipment alignment, pipe strain, dynamic behavior and measured as-left acceptance. A foundation may be structurally adequate yet produce unacceptable machine distortion; conversely a machine may align cold but move outside its applicable limit after piping, thermal stabilization or service loading.

Those remain with the qualified designer, OEM and competent installation personnel.

2. Allocation, boundaries and evidence hierarchy

3. Identify the package before specifying the foundation

Record the compressor package model, serial, revision, skid/baseplate assembly, driver type, coupling type, ancillary tanks, silencers, cooler arrangements, filled and operating mass, centre of gravity, lifting points, anchor layout and dynamic operating envelope. A portable package with isolation mounts does not share the same installation logic as a rigidly grouted industrial baseplate. Do not assume an OEM illustration applies to a later serial revision or to a package modified after delivery.

Establish whether the package has a common skid, separate driver and compressor pedestals, a fabricated baseplate with grout ports, pregrouted rails, spring isolators, elastomer pads or a manufacturer-designed modular foundation. Each architecture changes how support loads travel to the concrete and which movements are permitted. Record which dimensions are as-designed, as-shipped and as-found; never treat a catalogue drawing as a surveyed site drawing.

4. Site duty and machine dynamics

The civil design input must distinguish dead mass from operational, transient and maintenance loads. Relevant cases include starting torque, torsional pulsation, unbalanced reciprocating forces, pressure pulsation transmitted through piping, emergency stops, external nozzle loads, rotor run-up, resonance and adjacent machinery vibration. Actual excitation frequencies depend on machine speed, cylinder arrangement, firing order or rotating elements; a generic single “vibration allowance” cannot replace a dynamic assessment.

A foundation design review should document soil/geotechnical assumptions, slab thickness and reinforcement, embedment, load transfer, isolation strategy, drainage, access, seismic/wind requirements where applicable and allowable settlement or rotation. Civil acceptance must be signed by the responsible designer; an installation technician cannot infer structural adequacy from the absence of visible cracks.

5. Interface responsibility matrix

6. Concrete readiness and dimensional survey

Before grout placement, compare anchor positions, elevation, level, edge distances, embedment, grout envelope, equipment clearances and access with the latest approved drawings. Survey with a stated datum and instrument accuracy. Record the concrete condition: age, verified strength evidence, visible cracking, honeycombing, contamination, moisture condition, laitance removal and any repair. A surface may look clean while containing curing compounds or oil that undermine grout bond.

Surface preparation must follow the selected grout and structural specification. Roughness, saturation state, substrate temperature, standing water restrictions and bond requirements vary by product. Do not substitute a generic “roughen and wet” instruction for the product data sheet. Record any departures and obtain written disposition before placement.

7. Anchors: position, restraint and load path

Confirm anchor grade, diameter, length, projection, washer/plate configuration, embedment and orientation against the approved design. Cast-in and post-installed anchors have different qualification and installation requirements. For chemical anchors, hole cleaning, adhesive batch, substrate conditions and cure before loading are critical; for mechanical anchors, installation method and substrate condition are equally controlled. The applied torque does not by itself prove the correct preload or anchor capacity.

Anchors must restrain the equipment without forcing the baseplate into distortion. Tightening sequence and final torque/preload must come from the approved supplier procedure. Log the instrument identification and calibration, witness marks if specified, and post-grout recheck. Any cut, welded, bent, replaced or relocated anchor is a design change requiring engineering disposition.

8. Grout system selection and compatibility

Cementitious and polymer/epoxy grout systems differ in compressive behavior, thermal expansion, chemical resistance, creep, working time, thickness limitations, exotherm, vibration response and environmental sensitivity. Selection should be based on the machine/baseplate design, required load transfer, foundation exposure and manufacturer data rather than “higher strength is always better”. A stiff material may not be suitable where differential movement or vibration isolation is designed into the package.

Record grout product and revision, mixing ratio, batch traceability, storage history, substrate preparation, pour depth, ambient/substrate/material temperature, formwork, vent paths, working time, placement sequence and curing requirements. Field substitutions require compatibility review. The as-built grout interface must be accessible for void detection or verification where the approved method requires it.

9. Grout geometry, venting and void risk

The grout is intended to transfer specified baseplate reactions into the foundation. It cannot do so reliably if load-bearing pockets remain unsupported. Identify grout ports, vent holes, under-base obstructions, sleeves, anchor pockets, minimum/maximum permitted grout thickness, chamfers and edge termination. The placement method should avoid trapping air or creating cold joints, with the grout supplier’s instructions controlling actual execution.

A surface that appears filled does not demonstrate full underside contact. Evidence may include approved witness points, documented grout consumption versus theoretical envelope, inspection openings or an approved nondestructive check. Such checks have limitations and should not be described as conclusive without method validation. Distinguish cosmetic perimeter cracking from cracks that may indicate structural or bond distress.

10. Temporary supports, shims and leveling

Document the support philosophy: permanent steel shims, leveling screws removed after cure, chocks, soleplates, fitted wedges or proprietary mounts. Temporary leveling screws can create point loading if left bearing after grout cure contrary to the design. Conversely removing supports that the manufacturer intends to remain can destabilize alignment. Capture the as-found support state before any adjustment.

Survey elevations at defined points rather than reporting a single “level” number. A four-corner reading can reveal base twist; a straightedge can miss local foot contact conditions. Reference each measurement to a stable datum, record temperature and instrument uncertainty, and preserve the baseline for later pipe connection and thermal checks.

11. Soft foot: distinguish types and mechanisms

Soft foot means the mounting arrangement changes machine geometry when a hold-down is tightened or released. Common mechanisms include parallel gap, angular contact, induced soft foot from baseplate distortion, debris, damaged feet, inappropriate shim packs and pipe-induced strain. Soft-foot correction must follow the machine and coupling manufacturer’s measurement method and tolerance. An arbitrary shim addition can conceal an underlying structural or piping defect.

Document which foot was evaluated, fastening condition, gauge/laser method, repeatability, as-found movement and correction. Recheck after grout cure, anchor tightening, piping connection and relevant temperature stabilization. If a reading changes when an adjacent anchor is tightened, investigate baseplate flexure and load distribution rather than accepting an average value.

12. Alignment geometry and measurement references

Define whether the coupled train is horizontal, vertical, belt-driven or directly coupled; whether the OEM specifies offset/angular tolerances, rim/face readings, laser targets, thermal-growth offsets or shaft-position limits. Record measurement plane, coupling separation, shaft rotation direction, bracket sag or laser setup, runout, backlash, and compensation method. A number without its reference plane and temperature state is not a defensible alignment result.

Coupling misalignment may be expressed as offset at the coupling and angular slope, or as correction moves at front and rear feet. Those values are not interchangeable. Document the transformation, units and sign convention. Final acceptance uses the applicable OEM/coupling criteria at the specified operating state, not an internet table of “acceptable alignment”.

13. Worked example A — geometric correction from measured angular slope

Illustrative training case: a laser alignment instrument reports vertical offset of +0.12 mm at a defined coupling plane and angular slope of +0.08 mm per 100 mm. The movable machine front and rear feet are 250 mm and 850 mm from that plane. Under a simplified straight-line model, required relative corrections at those stations are 0.12 + (0.08/100 × 250) = 0.32 mm and 0.12 + (0.08/100 × 850) = 0.80 mm. Signs and direction depend on the actual instrument convention. These are geometric predictions, not permission to shim or a pass/fail criterion.

Before acting, confirm shaft reference, foot distances, instrument geometry, thermal targets, bolt-bound constraints and measurement repeatability. If a correction is mechanically impossible, inspect base twist, pipe loads or incorrect machine position. Record the as-left laser report and compare with the exact coupling and OEM acceptance limits.

14. Thermal growth and hot alignment

Cold alignment may intentionally be offset to achieve operating alignment after differential heating. For a simplified unconstrained member, ΔL = α × L × ΔT. For a steel support height of 0.65 m, assumed expansion coefficient 12 × 10⁻⁶/K and 45 K temperature rise, calculated free expansion is 0.351 mm. Real machinery may be restrained, have uneven temperatures or use different materials; this number is only a sensitivity estimate, not a machine-specific cold target.

Thermal movement should be derived from manufacturer targets or measured operating-state evidence. Record machine temperatures, process state, stabilization period, coupling condition and measurement technique. A hot check must follow an approved safe method; never expose personnel to rotating equipment or bypass guards for convenience.

15. Pipe strain and external connection effects

Suction, discharge, cooling, drain, exhaust and electrical connections can impose forces or moments on the package. Flexible connectors do not automatically eliminate loads; poor installation, unsupported hoses or excessive offsets may transmit substantial reactions. Record nozzle-load limits where provided and verify supports, guides, thermal loops and expansion joints against the approved piping design.

A controlled connected/disconnected alignment comparison may identify pipe-induced movement, but breaking pressure piping is a high-risk task requiring isolation, depressurization and a qualified method. Never recommend loosening live flanges. If connecting piping shifts machine alignment materially, correct the piping support/interface rather than pulling the machine into position with flange bolts.

16. Worked example B — piping connection shift

Illustrative case: repeated cold alignment measurements show 0.04 mm offset with the discharge spool absent and 0.23 mm after the spool is connected, at the same coupling plane and temperature. The observed change is 0.19 mm. This is evidence of an installation-induced shift, not proof that the flange alone caused it. Investigate support elevations, spool fit-up, anchor load redistribution and repeatability; document both configurations and obtain piping/OEM disposition before final acceptance.

17. Vibration commissioning and baseline interpretation

Vibration should be recorded with sensor location, axis, attachment, frequency band, machine speed, load, operating pressure, temperature and instrument settings. A single overall velocity figure without these conditions can obscure resonance, structural looseness, piping excitation or bearing-related phenomena. Trend spectral peaks with speed and operating state; avoid declaring a foundation defective solely from one elevated measurement.

A useful acceptance package compares pre- and post-installation signatures at equivalent duty. Look for changes at running speed, harmonics, subharmonics, vane/blade pass or reciprocating orders, and foundation-to-bearing transmission. Distinguish machine excitation from structural amplification using qualified diagnostic methods. OEM limits and applicable machine classification, not arbitrary web thresholds, control disposition.

18. Worked example C — suspected structural amplification

Illustrative case: measured RMS vibration at a compressor bearing is 3.2 mm/s at 1,480 rpm and at a nearby baseplate point is 2.8 mm/s, while a foundation point records 0.7 mm/s under the same load. A speed sweep shows a pronounced baseplate response near one operating speed. The pattern suggests a structural response worth investigating but does not establish resonance or an unsafe condition by itself. A qualified vibration specialist should review spectral data, operating modes, measurement uncertainty and design requirements.

19. Alignment measurement uncertainty and repeatability

Instrument accuracy is only one contributor. Mounting repeatability, temperature drift, shaft runout, bracket sag, coupling backlash, foundation settlement, operator technique and measurement-plane definitions can dominate field uncertainty. Record at least repeat measurements when the method requires them. If a result sits near a stated tolerance, quantify whether the uncertainty could change the pass/fail decision and seek OEM disposition rather than rounding toward acceptance.

Illustrative uncertainty illustration: a measured offset is 0.15 mm with an expanded uncertainty estimate of ±0.05 mm, while a hypothetical project-specific limit is 0.17 mm. The measurement interval overlaps the limit, so a simple “0.15 < 0.17” conclusion is not robust. Do not generalize this hypothetical limit to any real machine.

20. Worked example D — baseplate survey and differential elevation

Four surveyed reference points relative to a benchmark read +0.10, +0.12, +0.28 and +0.11 mm. The highest-to-lowest spread is 0.18 mm. This is a geometric descriptor only. It does not identify twist without point locations, load conditions and a fitted plane, nor does it determine acceptability without the approved tolerance. Record coordinates, datum, temperature, instrument and survey method; investigate the outlying point for shim/support condition and local deformation.

21. Grout defect investigation and repair boundary

Cracking, debonding, moisture ingress, chemical attack, loss of contact, grout edge crushing or anchor-pocket damage require defect characterization before repair. Document defect position, length/width/depth where measurable, load-bearing significance, exposure, photographs and progression. Repair choices—localized injection, replacement, regrouting, anchor remediation or design modification—must be selected by the responsible engineer and grout manufacturer. A cosmetic patch does not prove restoration of load transfer.

Retest after intervention using the approved criteria: support contact, anchor status, alignment, piping effect, vibration and any civil inspection. Preserve the defect and repair record with material batch, cure and witness evidence. Do not return equipment to duty merely because the visible crack has disappeared.

22. Worked example E — grout volume reconciliation

Illustrative under-base envelope: 1.8 m × 0.75 m × 0.045 m = 0.06075 m³, or 60.75 L before subtracting obstructions and adding edge forms, losses and anchor pockets. If field consumption is 48 L, the 12.75 L difference is not automatically a void: actual clearance, supports, geometry, waste and grout yield must be reconciled. A quantity discrepancy is a prompt for inspection, not proof of a hidden defect or acceptance.

23. Worked example F — thermal survey before acceptance

A fabricated skid has two measured support heights of 600 and 625 mm and measured operating temperature rises of 25 K and 40 K respectively. Using an illustrative steel coefficient 12 × 10⁻⁶/K, predicted unconstrained expansions are 0.180 and 0.300 mm. The difference is 0.120 mm. This calculation flags possible differential movement but cannot substitute for measured hot alignment because real support restraints, thermal gradients and stiffness govern the final geometry.

24. Installation Interface and Measurement Record — mandatory editable asset

Every row requires an evidence source and revision, input or measurement, method, result/status, named owner, unresolved item and disposition. Blank evidence is “not verified”, never “pass”. The following register is a developer-ready field inventory and may be converted into a controlled form with required-field validation.

25. Stage-gated installation workflow

25.1 Design freeze

Verify package reactions, support philosophy, foundation design, drawings and responsibilities. Gate status must be recorded as accepted, rejected or pending; missing evidence is pending. Any rework that affects alignment, anchors or load path requires targeted re-verification of dependent records.

25.2 Civil readiness

Inspect concrete, anchors, datum, dimensions, surface and access; sign off before grout. Gate status must be recorded as accepted, rejected or pending; missing evidence is pending. Any rework that affects alignment, anchors or load path requires targeted re-verification of dependent records.

25.3 Baseplate setup

Record supports, level, soft foot, anchor condition and pregrout geometry. Gate status must be recorded as accepted, rejected or pending; missing evidence is pending. Any rework that affects alignment, anchors or load path requires targeted re-verification of dependent records.

25.4 Grout execution

Record material, mixing, conditions, placement, vents, cure and nonconformities. Gate status must be recorded as accepted, rejected or pending; missing evidence is pending. Any rework that affects alignment, anchors or load path requires targeted re-verification of dependent records.

25.5 Mechanical assembly

Verify fasteners, shaft reference, cold alignment and piping interface. Gate status must be recorded as accepted, rejected or pending; missing evidence is pending. Any rework that affects alignment, anchors or load path requires targeted re-verification of dependent records.

25.6 Dynamic commissioning

Record controlled duty, vibration, temperatures, thermal targets and changes. Gate status must be recorded as accepted, rejected or pending; missing evidence is pending. Any rework that affects alignment, anchors or load path requires targeted re-verification of dependent records.

25.7 Final release

Close deviations, compile as-built evidence and obtain named OEM/civil/mechanical approvals. Gate status must be recorded as accepted, rejected or pending; missing evidence is pending. Any rework that affects alignment, anchors or load path requires targeted re-verification of dependent records.

26. Failure modes and diagnostic discrimination

27. Engineering hold points and prohibited assumptions

Hold work or return-to-service disposition when the machine serial cannot be reconciled to the approved installation drawings; foundation capacity or anchor design is unapproved; grout batch or installation conditions are untraceable where required; a support or anchor has been modified without authorization; alignment criteria are missing; measured deviations exceed limits or overlap uncertainty; significant pipe strain remains unresolved; or commissioning vibration shows an unexplained adverse change. These are evidence holds, not universal operating limits.

Do not prescribe a generic grout compressive strength, anchor torque, alignment tolerance, permissible nozzle load, vibration threshold, cure duration, concrete strength or foundation natural frequency. Each depends on equipment design, materials, jurisdiction and applicable manufacturer instructions.

28. FAQs — AEO and AI-search answer contract

Can a compressor be installed directly on a concrete floor?

Only if the specific OEM package design and approved civil assessment allow that mounting configuration. Skid isolation and anchoring requirements differ.

Is grout always required?

No. Some packages use proprietary isolation or pre-engineered supports; others require engineered grout contact.

Does a level baseplate prove alignment?

No. Machine foot geometry, coupling reference, piping strain and thermal movement must also be verified.

What is soft foot?

A mounting condition in which tightening or releasing a machine foot changes its supported geometry.

Can alignment be accepted using generic tolerances?

No. Apply the serial-applicable OEM and coupling limits for the specified duty and temperature.

Why does alignment change when piping is connected?

External forces, support offsets, thermal mismatch or skid movement may distort the machine interface.

Does grout cracking always require replacement?

No. The significance depends on location, progression, contact, load path and approved engineering evaluation.

Can an anchor be relocated in the field?

Only under an approved engineering change addressing load path, embedment and foundation capacity.

What should be recorded during commissioning?

As-left alignment, support/anchor status, vibration with duty, temperatures, outstanding deviations and signoffs.

What makes the evidence register complete?

Exact asset and revision, sourced limits, method, calibrated readings, results, owners and closed or formally assigned uncertainties.

What is the difference between cold and hot alignment?

Cold is measured before operating thermal stabilization; hot or compensated targets account for the relevant operating state.

Who approves return to service?

The named competent owner(s) under the project’s applicable OEM, civil, mechanical and legal responsibilities.