A Guide to Low Temperature Studies for Process Integrity

1. The Risk: Brittle Fracture

The primary purpose of a low-temperature study is to ensure the mechanical integrity of equipment and piping by preventing brittle fracture. This catastrophic, rapid failure of (typically) carbon steel occurs when three conditions are met simultaneously:

1. A material defect exists (e.g., a small crack or inclusion from manufacturing).
2. A high stress is applied (from internal pressure, thermal shock, or even residual welding stress).
3. The material has a lack of toughness, which for carbon steel is a direct result of being cooled below its design minimum temperature.

A catastrophic failure at a gas plant in 1998 was a direct result of liquid hydrocarbons cooling a heat exchanger to approximately -48°C, far below its design temperature. When warm liquid was reintroduced, the resulting thermal shock and high stress caused the embrittled steel to fail, leading to a massive release and explosion. This guide outlines the methodology to prevent such events.

2. When to Perform a Low Temperature Study

A low-temperature analysis must consider all scenarios—normal, abnormal, and transient—that could chill process equipment. A systematic review of the following potential causes is the first step.

3. The Overall Methodology: Screening to Detailed Analysis

A tiered approach is used to focus engineering effort where it is most needed.

1. Identify Scenarios: Using the table above, identify all credible low-temperature scenarios for each system.
2. Initial Screening: For each scenario, perform a simple, conservative calculation (e.g., a simple isenthalpic flash in a process simulator).
3. Assess Screening Results:
   • If the predicted temperature is well above the material's minimum design temperature (e.g., > 0°C for a Carbon Steel system rated to -29°C), no further analysis may be needed.
   • If the predicted temperature approaches or falls below the material's limit, a more rigorous, detailed analysis is mandatory.
4. Detailed Analysis: Perform a rigorous, time-dependent (dynamic) simulation that accounts for heat transfer, thermodynamics, and system geometry.
5. Specify Materials: The final, lowest predicted metal temperature (with a design margin) is used to confirm material selection.

4. Key Thermodynamic Concepts

"Isentropic" vs. "Isenthalpic" Expansion

Understanding these two terms is critical for identifying the correct "cold" location.

• "Isentropic-like" Expansion (In the System): This is gas expansion *within* a vessel or pipe that is being depressured. The gas does work as it expands, causing the bulk fluid temperature to drop. This determines the coldness of the upstream equipment.
• "Isenthalpic" (Joule-Thomson) Expansion (Across a Valve): This is a constant enthalpy pressure drop across a restriction (like a control valve, orifice, or PSV). This determines the coldness of the downstream piping.

The Impact of Kinetic Energy (High Velocity)

In a standard "isenthalpic" flash, we assume H_in = H_out. However, in a high-velocity line (like a blowdown tailpipe), the energy balance is actually H_in = H_out + KE_out, where KE is Kinetic Energy.

This means that some of the fluid's enthalpy (heat) is converted into kinetic energy (velocity). This makes the fluid's static temperature even colder than a simple isenthalpic flash would predict. This effect is most significant immediately downstream of a restriction where velocity is highest (approaching Mach 1).

• During Depressuring: The gas is extremely cold at the start, but the high-velocity flow is short-lived. The pipe wall has "thermal inertia" and doesn't have time to cool down completely before the flow (and thus the kinetic energy effect) subsides. The net effect on the *minimum metal temperature* is often small.
• During Pressure-Up: The high-velocity cold gas flow is sustained, potentially for a long time. The pipe wall *will* be chilled to this very low temperature. This is often the governing case for the tailpipe material.

5. Dynamic Simulation: Best Practices

Lumped vs. Distributed Models

A lumped model (one big pot) is acceptable only for screening a single, simple vessel. A distributed model, which models the vessel, inlet pipe, blowdown valve, and segments of the tailpipe as separate, connected units, is required for an accurate analysis.

Modeling Initial Conditions

The worst-case starting point is often not the hot, normal operating condition. The most conservative case is frequently after a shutdown, where the system has cooled to ambient temperature but remains at high pressure (e.g., due to a leaking valve). This "isochoric" (constant volume) cooling results in a much lower starting temperature before the depressuring even begins.

The "Worst-Case" Depressuring Duration

It is incorrect to assume that the fastest depressuring (e.g., a 15-minute blowdown) is always the worst case for metal temperature. A "worst-case" duration often exists:

• Too Fast: The gas gets extremely cold, but the event is over so quickly that the metal (which has thermal inertia) doesn't have time to cool down.
• Too Slow: The depressuring is so gentle that heat transfer from the environment keeps the pipe and fluid warm.
• "Just Right" (Worst Case): A specific duration (e.g., 30-60 minutes) may exist where the chilling is sustained long enough to overcome the pipe's thermal inertia, resulting in the lowest minimum metal temperature. A sensitivity analysis on duration is recommended.

Handling Liquids and Water

• Hydrocarbons: The most conservative case for a vessel is typically the Low-Low Liquid Level (LLLL). Less liquid means less thermal mass ("heat sink") available to keep the system warm.
• Water: The presence of free water or hydrates complicates the simulation. As water freezes, it releases a significant amount of latent heat, which can "arrest" the temperature drop around 0°C. However, this is difficult to model reliably.

6. Practical Design for Low Temperatures

Blowdown Piping and Tailpipe Configuration

The coldest temperatures almost always occur in the blowdown/flare piping immediately downstream of the blowdown valve or restriction orifice (RO). The goal is to minimize the length of expensive, low-temperature (e.g., Stainless Steel) piping.

7. The Result: Specifying the Minimum Design Temperature (MDMT)

The final output of the study is the MDMT for each piece of equipment. This is the lowest temperature the metal is expected to reach, and it dictates the material selection.

How to Define the MDMT

The MDMT is the lowest of the following:

• The minimum ambient temperature for the site.
• The minimum normal operating temperature, minus a design margin (e.g., -5°C).
• The lowest predicted temperature from a dynamic simulation, plus a design margin (e.g., +/ -5°C or 10°C, depending on confidence in the model).

Exception: If a liquid is at its boiling point (e.g., a de-ethanizer bottom), the boiling point is a thermodynamic floor. No margin is required, as the liquid *cannot* get colder. The vessel wall temperature will be equal to this liquid temperature.