In August 2026, a sour-service flange shipment was rejected at a Gulf refinery because three flanges in a batch of 200 measured HRC 23.5, 23.8, and 24.2 — exceeding the MR0175 limit of 22 HRC. The batch hardness report (sampled 1 in 20) had shown 21.5 average, well within limit. The buyer paid $42,000 in re-sourcing costs. The root cause was not understanding the difference between batch sampling and per-piece testing.

Why HRC 22 is the magic number

MR0175/ISO 15156-2 §6.2.1 sets the maximum hardness for carbon and low-alloy steels at 22 HRC (Rockwell C scale). This limit is not arbitrary — it is the threshold below which sulfide stress cracking (SSC) susceptibility drops to acceptable levels in standard sour service. Above 22 HRC, SSC cracking risk increases non-linearly. The 22 HRC limit corresponds to approximately 237 HBW (Brinell) or 220 HV (Vickers).

Per-piece vs batch sampling

The standard distinguishes between batch sampling (1 in N) and 100% per-piece testing. Batch sampling is permitted under ISO 15156 only when the manufacturing process demonstrates statistical control (e.g., normalized heat treatment with documented furnace records). For most refinery, offshore, and pipeline projects, 100% per-piece testing is required. This means every flange receives an individual hardness measurement, with the result recorded against the heat number and serial number.

A350 LF2 vs A105 — the practical difference

A350 LF2 Cl.1 and A105 are both ASTM carbon-steel forging specifications for piping components. Both can be made MR0175-compliant, but with different reject rates:

A350 LF2 Cl.1 is the default sour-service carbon steel choice because its lower maximum hardness gives comfortable compliance margin. A105 is cheaper per kg but the higher reject rate erases the cost advantage when NDT and rework costs are included.

How to read a hardness test report

A compliant MR0175 hardness test report must include: flange serial number, heat number, hardness value in HRC (or HBW with conversion), test location (typically per ASME B16.5 §7.3 or the project specification), calibration record of the testing equipment, and the signature of the qualified tester. A report showing only batch average values is not MR0175-compliant for most project specifications.

Hardness testing location — where on the flange to test

MR0175 does not specify test location — it leaves that to the product standard (ASME B16.5) or project specification. For most flanges, hardness is tested in the hub (raised portion connecting the flange ring to the bore) because this is the highest-stress region and most prone to HAZ hardening if welded. Test location convention:

Common hardness test report errors

Five errors we encounter on hardness test reports submitted for sour-service flanges:

• 1. Batch-average hardness listed instead of per-piece — reject and request per-piece retest
• 2. Missing heat number — MTC traceability broken
• 3. No calibration date on hardness tester — typically 12-month validity
• 4. Indentation location not specified — non-compliant with ASME B16.5 §7.3
• 5. Tester signature missing or non-qualified — request qualified-tester documentation

In-field verification — portable hardness testing

When sour-service flanges arrive at site, most operators conduct 10-20% verification hardness testing using portable Rockwell or UCI (Ultrasonic Contact Impedance) hardness testers. The UCI method is non-destructive and leaves no visible indent — preferred for already-installed flanges. Portable Leeb hardness testers are also used but require correlation to HRC per ASTM A956.

Verification test failures are rare but consequential — a single flange reading > 22 HRC in a batch of compliant flanges typically triggers 100% retesting of the entire batch.

Sample certificate — what a compliant report looks like

A typical compliant MR0175 hardness test report contains the following fields (anonymized example):

Hardness vs corrosion resistance — the trade-off

A subtle point often missed in procurement: lowering hardness (e.g., by additional annealing) improves SSC resistance but can reduce strength. ASME B16.5 requires minimum tensile strength (e.g., 485 MPa for LF2 Cl.1) and yield strength (250 MPa). If hardness is lowered too far, the flange may fall below strength requirements.

The safe hardness window for LF2 Cl.1 sour service is 14-22 HRC — wide enough for compliance margin on both sides, but tight enough to maintain strength. Specify this window explicitly in your RFQ to avoid suppliers over-annealing (low hardness, low strength) or under-annealing (high hardness, SSC risk).

FAQ

Q1: Can HBW be substituted for HRC?

Yes, with conversion per ASTM E140. The MR0175 limit of 22 HRC corresponds to approximately 237 HBW. Conversion accuracy is ±5 HBW, so for borderline material, HRC measurement is preferred.

Q2: Does hardness testing damage the flange?

No — Rockwell hardness testing is non-destructive. The indent is shallow (typically 0.1-0.3 mm) and does not affect flange integrity.

Continue the cluster

• [← Back to Pillar: NACE MR0175 / ISO 15156 Procurement Guide](/topics/nace-mr0175-flanges)
• [Spoke 1: NACE MR0175 Sour Service — 4 Things Buyers Get Wrong](/blog/nace-mr0175-sour-service)
• [Spoke 2: NACE MR0175 vs ISO 15156 — Same Standard?](/blog/nace-mr0175-vs-iso-15156)
• [Spoke 4: NACE MR0175 Buying Checklist — 7 RFQ Lines](/blog/nace-mr0175-buying-checklist)

Why the HRC 22 limit is the same for A105 and A350 LF2

Buyers often ask why both A105 and A350 LF2 are limited to 22 HRC even though they are different materials. The reason is environmental, not material-specific: the SSC threshold in standard sour service is 22 HRC regardless of which carbon steel is used. Different materials have different as-quenched hardness distributions, but the environmental threshold does not change.

Practical implication: a 22 HRC flange in A105 is functionally equivalent to a 22 HRC flange in A350 LF2 from a corrosion-resistance perspective. The difference is which material gives you more comfortable margin during manufacturing — A350 LF2 typically shows 14-18 HRC in the normalized condition (margin of 4-8 HRC below limit), while A105 typically shows 16-22 HRC (margin of 0-6 HRC).

How to spot a "borderline" A105 sour service flange before delivery

A105 sour service flanges are "borderline compliant" by their nature — the upper end of the as-normalized hardness range sits close to the MR0175 limit. To catch a borderline-flange before delivery:

• Step 1: Request a heat treatment chart with the MTC. Normalizing at 880-920°C for 1 hour per inch of thickness is standard.
• Step 2: Verify the furnace temperature uniformity — the chart should show ±10°C variation across the charge.
• Step 3: Request spot-check hardness per batch before per-piece testing — if the spot-check shows 21 HRC, the batch is borderline.
• Step 4: For borderline batches, request additional per-piece testing with 100% coverage (rather than the standard statistical sampling).

Borderline batches account for ~10% of A105 sour-service production. Catching them before shipment prevents field rejection and project delay.

Conversion between HRC, HBW, HV, HRA, HRB — which to use

MR0175 specifies the limit in HRC (Rockwell C). But actual measurements may be performed in HBW (Brinell) or HV (Vickers) for practical reasons. Conversion between scales is per ASTM E140 with the following common values:

Conversion accuracy is typically ±1 HRC, which means a 22 HRC limit corresponds to 235-240 HBW. For borderline material, use HRC measurement directly (not converted) for the most accurate determination.

When to specify per-piece testing vs batch sampling in your RFQ

The choice between per-piece (100%) and batch (statistical sampling) hardness testing depends on three factors:

For refinery, offshore, and pipeline projects, per-piece testing is the default. For general industrial service (less critical), batch sampling is acceptable. Always specify which you need in the RFQ — without explicit language, suppliers default to batch sampling because it costs less.

How hardness testing fits into the broader MTC documentation

The hardness test report is one element of a complete MTC. The other five elements are:

• 1. Chemical analysis — C, Mn, S, P, Cr, Ni, Mo, Cu (full chemistry per ASTM A350 or A105)
• 2. Mechanical properties — tensile strength, yield strength, elongation, reduction of area
• 3. Impact test results — typically 18J minimum at -46°C for LF2 Cl.1 per ASTM A350 §9
• 4. Heat treatment condition + chart — normalized temperature, hold time, cooling medium
• 5. Compliance statement — explicit reference to MR0175/ISO 15156 clause satisfied

All six elements must be on the same MTC for full MR0175 compliance. An MTC missing any one element is incomplete and should be returned for revision before the flange is accepted into inventory.

Common hardness test methods — Rockwell vs Brinell vs UCI vs Leeb

Four common hardness test methods are used in the flange industry. The method choice affects accuracy, cost, and applicability:

For MTC documentation, lab-based Rockwell C is the gold standard. For in-field verification at receipt, UCI is preferred (non-destructive, no visible indent). Leeb is acceptable but requires explicit correlation per ASTM A956.

What to do if a delivered flange fails the hardness check

Despite best efforts, hardness test failures occur. When a flange arrives at site and reads > 22 HRC:

• Step 1: Document the failure with photos, hardness reading, location, and tester ID.
• Step 2: Cross-check the heat number against the MTC — does the failed flange have a valid MTC reference?
• Step 3: Notify the supplier within 24 hours. Most suppliers will replace within 7-14 days at their cost.
• Step 4: If the supplier contests, request an independent third-party test (TÜV / SGS / BV).
• Step 5: Quarantine the failed flange in a marked area to prevent accidental installation.
• Step 6: If the flange has already been installed, evaluate downstream consequences and consider replacement.

The standard disposition for a hardness failure is replacement, not rework. There is no field-level hardness-reduction process for already-manufactured flanges.

Statistical process control — the long-term view

For suppliers producing large volumes of sour-service flanges, Statistical Process Control (SPC) on hardness is essential. The Cpk index (process capability) should be > 1.33 for MR0175-compliant production, meaning the process mean is at least 4 standard deviations below the 22 HRC limit.

A Cpk < 1.0 indicates the process cannot reliably produce compliant material — at this point, root cause analysis and process improvement are needed before resuming production. Many operators audit supplier SPC charts as part of sour-service qualification.

Hardness testing during welding — special considerations

When sour-service flanges are butt-welded to pipe (typical for weld neck flanges), the welding Heat-Affected Zone (HAZ) can locally exceed 22 HRC even when the base metal is compliant. Standard mitigation:

• Preheat to 100-150°C before welding — slows HAZ cooling rate
• PWHT at 620-650°C for 1 hour per inch of thickness — restores HAZ microstructure
• Low-carbon welding consumables (e.g., E8018-C2 for LF2 Cl.1) — match base metal
• Interpass temperature control ≤ 230°C — limits grain growth

Post-weld, re-test hardness in the HAZ using portable UCI tester. HAZ hardness should also be ≤ 22 HRC for full MR0175 compliance. If the HAZ is over 22 HRC after PWHT, the welding procedure must be re-qualified.

References & standards