Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ASME STP-PT-079-2016 — Local Heating of Piping: Thermal Analysis
📐 ASME STP-PT-079-2016 is a technical report sponsored by the ASME Pressure Technology Codes & Standards committee and ASME ST-LLC, prepared by Quest Integrity USA, LLC and issued in June 2016. It performs a systematic thermal analysis of local heating of piping — including local post-weld heat treatment (PWHT), preheat, and insulation practices — quantifying the temperature distribution and the inner-to-outer-wall temperature difference (ΔT) during heating, and aligning the results with the requirements of AWS D10.10 (Standard for Local Heating of Welds in Piping and Tubing) to support thermal-stress and temperature-control assessments in the ASME code system.
💡 Local heating (instead of furnace treatment of the whole pipe) is the standard practice for large-bore piping and in-service line repair. The core trade-off: the narrower the heating band, the more energy-efficient and convenient, but the larger the through-wall ΔT and axial gradient — and with them the thermal-stress risk. This report quantifies exactly that trade-off.
1. Background and Study Matrix 📌
In post-weld heat treatment and preheat of thick-wall piping, site constraints usually require electric heating blankets wrapped around the pipe to form a local heating band. Local heating creates a radial temperature difference across the cross-section (the outer wall is heated directly, the inner wall lags behind) plus an axial gradient, generating thermal stresses. For martensitic creep-resistant steels such as P91, temperature control and cooling rate directly govern hardness and tempered microstructure, and excessive ΔT can cause cracking.
The report analyzes five nominal pipe sizes — 6, 10, 14, 24, and 30 inches — each covering Schedule 80, 120, and 160 wall series, all in P91 material, forming a systematic computational matrix.
| Parameter Category | Range | Purpose |
|---|---|---|
| Nominal pipe size | 6, 10, 14, 24, 30 in. | Small to large bore |
| Wall series | Schedule 80 / 120 / 160 | Different wall-thickness gradients |
| Material | P91 (9Cr-1Mo-V) | Creep-grade steel, temperature-sensitive |
| Heating method | Circumferential local band (blankets) | Field PWHT/preheat |
| Analysis method | CFD thermal analysis (radiation, contact resistance) | Predict temperature field and ΔT |
2. Methodology: Calibration and Prediction Models 🔧
The thermal analysis proceeds in two stages:
- Physics modeling: solid conduction, surface radiation, and thermal contact resistance between the heating blanket and the pipe (two-body thermal contact) are included; the blanket is modeled at its real power density, using half-/full-symmetry reduced domains.
- Calibration model cases: measured temperatures (e.g., centerline inner/outer wall) are used to reverse-engineer parameters that are hard to measure directly — such as the thermal contact resistance (e.g., 0.0037 m²·K/W) — ensuring model/test agreement.
- Prediction model cases: the calibrated model then predicts temperature distributions for each size/wall combination, outputting the through-wall ΔT, circumferential inner-wall profiles, and other characteristic quantities.
- AWS D10.10 alignment: results are organized into engineering curves matching the AWS D10.10 soak-band (SB) ΔT requirements, with power-law fit coefficients so the CFD results can be extrapolated to uncomputed cases.
⚠️ The results clearly show that for the same blanket setting, ΔT varies markedly with outside diameter and wall ratio (t/OD): the ΔT of large-bore thin-wall pipe is relatively controllable, while a small-bore (6 in.) thick-wall (Schedule 160) combination shows significantly larger ΔT. There is no single “one-size-fits-all” heating setting.
3. Key Results and Engineering Curves 📊
The report presents CFD ΔT results for each size/wall combination and compares them with the AWS D10.10 ΔT limits for local heating bands (HB/SB). Main findings:
- For 6–30 in. Schedule 80/120/160 P91 pipe, the band inner-to-outer-wall ΔT follows a stable power-law relationship with OD and wall ratio t/OD, allowing rapid estimation from fit coefficients.
- For a given ΔT target (e.g., an AWS D10.10 limit), the required band width and power configuration can be derived, guiding field execution.
- The calibration models confirm that, once contact resistance is included, predicted centerline temperatures agree closely with measurements, validating the methodology.
🚨 Martensitic steels such as P91 are extremely sensitive to local heating: improper band settings that cause large through-wall ΔT and uncontrolled cooling rates can induce reheat cracking during PWHT or incomplete transformation of retained austenite, resulting in weld-joint hardness violations and early failure.
4. Engineering Practice Essentials 🛠️
- 🔍 Heating-band design: determine soak-band (SB) and heat-blanket (HB) widths per AWS D10.10 and company procedures by pipe size and wall thickness; widen the band where possible to reduce ΔT.
- 📊 Thermocouple placement: place thermocouples at inner and outer wall temperature-control points to monitor ΔT; add multiple axial points on thick-wall pipe to capture the gradient.
- 🔧 Power and heating/cooling rates: control ramp rates by material and wall (e.g., P91 commonly within 200°C/h with specified upper limits) and use the report’s curves to estimate power needs for the target.
- 📋 Records and verification: record actual ΔT and curves during field PWHT and compare with predictions; if deviations are significant, re-check blanket contact and coverage and recompute if necessary.
5. Frequently Asked Questions ❓
❓ Q: Why does local heating create an inner-outer wall temperature difference?
A: The outer wall is heated directly by blanket radiation and conduction, while the inner wall receives heat only through wall conduction. Thick walls have high thermal resistance and surface heat loss makes the inner wall lag, producing a through-wall ΔT — and the larger the ΔT, the larger the thermal stress.
A: The outer wall is heated directly by blanket radiation and conduction, while the inner wall receives heat only through wall conduction. Thick walls have high thermal resistance and surface heat loss makes the inner wall lag, producing a through-wall ΔT — and the larger the ΔT, the larger the thermal stress.
❓ Q: What is the relationship between this report and AWS D10.10?
A: AWS D10.10 states the process requirements for local preheat and PWHT (band widths, ΔT and rate limits); this report quantitatively computes the actual ΔT for various pipe sizes and walls via CFD, converting the code’s “ΔT limit” into a “heating configuration requirement.” They complement each other.
A: AWS D10.10 states the process requirements for local preheat and PWHT (band widths, ΔT and rate limits); this report quantitatively computes the actual ΔT for various pipe sizes and walls via CFD, converting the code’s “ΔT limit” into a “heating configuration requirement.” They complement each other.
❓ Q: Why was P91 chosen as the study material?
A: P91 (9Cr-1Mo-V) is the most widely used creep-grade steel in power plants and high-temperature piping, with strict PWHT temperature-window and cooling-rate requirements; it is the most demanding typical material for local-heating control, and the results also inform other Cr-Mo steels.
A: P91 (9Cr-1Mo-V) is the most widely used creep-grade steel in power plants and high-temperature piping, with strict PWHT temperature-window and cooling-rate requirements; it is the most demanding typical material for local-heating control, and the results also inform other Cr-Mo steels.
🔍 Q: Can the on-site ΔT directly use this report’s curves?
A: The curves are based on typical blanket configurations and P91 material properties and are suitable for preliminary estimation. If the actual configuration (blanket specs, coverage, insulation) differs, calibrate with the report’s method first or re-run a similar CFD model.
A: The curves are based on typical blanket configurations and P91 material properties and are suitable for preliminary estimation. If the actual configuration (blanket specs, coverage, insulation) differs, calibrate with the report’s method first or re-run a similar CFD model.
