November 2025

Special Focus: Process Controls, Instrumentation and Automation

Tales of innovative approaches for skin metal temperatures

IP: 10.3.138.59

In the oil, gas and chemical industries, industrial heaters are used to transfer energy to the various processes that distillate, crack or convert the product into value-added products. To accomplish this exercise, heater tubes are operated at high temperatures and are exposed to damaging mechanisms such as oxidation, carburization, sulfidation, metallurgical embrittlement and/or creep.¹ Heater tubes are designed to withstand certain temperatures and pressures, and to provide safe and reliable operations, these parameters must be strictly followed.  

One challenge encountered in the industry is the difficulty to complete a precise, accurate measurement of tube metal temperature. This temperature must be monitored since it is possible to follow the integrity operating window² of this component to obtain a reliable measurement of degradation related to corrosion activity or creep, which are high temperature long-term damage mechanisms. Using industry recommended practice API-579 Part 10, tube metal temperature is also used to calculate remaining life.³  

Typically, heater tubes are monitored with thermography and/or skin thermocouple instruments.⁴ If these instruments are over measuring temperature, it will cause loss of production, and operations will lose confidence in these instruments’ measurements. Conversely, if the sensor is underestimating temperature, tubes can be exposed to a more aggressive environment and can increase operating risk. FIG. 1 provides a review of these situations. 

FIG. 1. The impacts of over/under temperature measurements at an operating site. 

In a collaborative effort to enhance measurement accuracy and reliability, the authors’ companies conducted a series of comprehensive tests, beginning at the co-author’s company^a’s research and development (R&D) center. The primary goal of this initial phase was to evaluate the accuracy of instrumentation while assessing how design modifications and installation challenges could affect performance. Building on the insights gained, the next phase of testing took place at the Suncor refinery, allowing for additional real-world validation under operational conditions. Suncor Energy produces synthetic crude from oil sands crude and refines it into various products at four refineries in Canada and the U.S. Following a year of field data collection, a thorough analysis was performed to understand the long-term impacts of these factors on measurement accuracy and overall system performance. 

History. In 2019, the co-author’s company^a expanded its R&D capabilities with the construction of a dedicated test facility in Houston, Texas (U.S.). This state-of-the-art facility enabled the organization to explore, test and refine innovative temperature measurement technologies under controlled conditions. As a result of this investment in R&D, the company successfully developed and patented several advanced tubeskin temperature measurement solutions.^5,6  

Meanwhile, Suncor had been utilizing various designs and construction methods for tubeskin temperature sensors across its operations. However, uncertainties remained regarding the performance and accuracy of these different sensor configurations. To address these concerns, Suncor sought a more in-depth understanding of how sensor design and installation variables impacted measurement reliability. This shared interest in improving sensor performance formed the foundation for a collaborative testing initiative between the authors’ companies. 

Test setup. The test setup at the co-author’s company’s Houston, Texas (U.S.) R&D center was designed to rigorously evaluate temperature measurement technologies under realistic and controlled conditions. The facility, spanning > 6,000 ft2 (560 m2), was engineered and constructed in accordance with ASME and API code requirements, ensuring adherence to industry standards. 7 Equipped with advanced data analysis capabilities, the center provided a robust environment for detailed performance assessment. Central to the testing was a custom-built furnace system featuring horizontal tube configurations. The furnace utilized a high-temperature heat transfer fluid, with a total capacity of 1,500 gal (5,600 l) and the output of a 9.7-MMBtu/hr (2,842-kW) ultra-low nitrous oxide (NOx) burner. It also offered the flexibility to adjust heat flux between 7 kBtu/h·ft² and 10 kBtu/h·ft² (22 kW/m2 and 32 kW/m2) and modify various conditions, allowing for comprehensive simulation of real-world operating environments. To further enhance the accuracy and depth of the evaluation, the co-author’s company incorporated infrared (IR) camera analysis and computational fluid dynamics (CFD) modeling as part of its testing methodology. FIGS. 2 and 3 represent the installation skin test setup that was completed.  

FIG. 2. Test setup skin installation in the heater radiant section. 

FIG. 3. Test setup skin installation in the heater convection section. 

Various technologies of skin thermocouples were installed in the furnace to complete current and past operation that Suncor experimented. In these installations, one sensor was deliberately welded incorrectly to quantify the impact of an improper installation. The sensors were also tested by x-ray to validate how they were installed. FIG. 4 shows the end view of the product documenting the improperly welded sensor. The welding parameters were selected to have approximately 1/8 in. (3 mm) of lack of fusion. Other design iterations using various shield geometry were also evaluated. The dimensions of the shield and amount of insulation were adapted to understand their importance. 

FIG. 4. Example weld showing lack of fusion (gap), denoted by the white arrow. 

The next step of testing was conducted at the Suncor site. The heater selected was a semi-regen catalytic reformer. This heater is comprised of four (A1-A2-B-C) different heating cells, and the design of each heat cell was built with an arbor configuration. This heater was selected because it is the biggest in the plant. It has a capacity of 415 MMBtu/hr (121,624 kW), which is enough energy to heat 40,000 houses. The cells that were selected to complete the test were A2 and B, where the highest heat flux was present. Cell A2 has a heat flux that varies between 22 kBtu/h·ft² and 27 kBtu/h·ft² (69 kW/m2 and 85 kW/m2), and Cell B varies between 30 kBtu/h·ft² and 33 kBtu/h·ft² (95 kW/m2 and 104 kW/m2).   

There are three rows of burners that are horizontally oriented. The infield testing setup was installed at the highest temperature and on the hottest tube. FIG. 5 provides a sketch of the heater coil configuration and location of installation. 

FIG. 5. Catalytic reformer heater sketch. The location of each skin is denoted in red. 

FIG. 6 shows the testing setup installation in the industrial heater. The biggest challenge of these tests was to complete the installation during a maintenance turnaround (TAR), where execution time is always limited due to the number of activities within the same location. In addition, this installation is like a laboratory test installation that is fragile. Continued good communication between the TAR execution team and the testing team was vital to make this experiment a success.   

FIG. 6. Testing setup in the catalytic reformer. 

During the testing, skin measurements were recorded to the site’s data historian. IR thermography was used to monitor the measurements, and skin reference sensors, designed by the co-author’s company, were installed to validate sensor readings against the thermographic data.  

Results. The comprehensive testing conducted at the co-author’s company’s R&D center evaluated multiple tubeskin thermocouple designs under controlled conditions, with a strong focus on understanding how inaccurate readings—whether high or low—can directly impact tube life and overall furnace safety. Each design was validated using CFD to simulate realistic heat transfer behavior. One key finding was the challenge of achieving consistently high-quality welds during sensor installation. As previously mentioned, a specific test involving a deliberately poor weld illustrated how improper attachment can significantly distort temperature readings, leading to potential underestimation/overestimation of actual tube temperatures. The effects of improper welding can lead to errors of nearly 30°F (16°C) or higher in extreme cases. While the installation itself has an impact, the sensor type exerts an even more significant influence on overall performance. Testing with alternative tubeskin sensor designs in the industry has demonstrated even greater accuracy deviations, exceeding ±80°F (±44°C). The additional tests using a shield verified that the temperature measurement could be underestimated in the range of 80°F–150°F (44°C–66°C). It must be noted that with the various designs tested, as the heat flux was adjusted, there was a correlation to the deviation of the sensors to varying degrees.  

Following lab validation, selected designs were installed in a fired heater at a Suncor refinery for in-service testing. Although heat flux in the refinery heater varied, it was carefully considered to replicate operational conditions. Sensor accuracy was confirmed through both reference sensor comparisons and IR imaging. To reduce variability in field installations, the authors’ companies implemented a strict protocol for the co-author’s company’s proprietary tubeskin thermocouple sensor technology^b, which included hands-on welder training using dummy tubeskins, followed by destructive testing of the installed sensors to ensure consistency and quality. After 1 yr of operation, data analysis confirmed that proper sensor design and installation led to more reliable and stable temperature measurements. 

These results highlighted the importance of including safety factors in standard operating procedures when interpreting tubeskin readings. Rather than adjusting the sensor output, it is often safer and more practical to revise alarm setpoints based on validated application data. Accurate tubeskin measurement is essential for furnace performance evaluation, especially when placed in critical areas such as outlet tubes, burner zones prone to flame impingement or regions where coking is likely to occur. Over time, sensor accuracy may drift, which is why periodic validation using IR imaging is recommended. Ultimately, precise and well-installed tubeskin sensors play a key role in protecting tube integrity, optimizing feed rates and ensuring safe, efficient reactor operation. 

In this case, the Suncor site used the results of this study to correct past temperature measurements to complete remaining life assessments using fitness-for-service evaluations. When creep assessment is completed, the precision of the temperature is very important. A fluctuation of 25°F (13°C) can affect the remaining life by a factor of two.^1,8,9 

Takeaways. Accurate tubeskin temperature measurement is essential for the safe and economical operation of fired heaters. Inaccurate readings can have serious consequences, with many industrial incidents recorded to highlight these concerns. When tubeskin temperatures are overestimated, it can unnecessarily limit unit production due to overly conservative operating constraints. High readings can also skew integrity calculations, potentially indicating accelerated creep damage, leading to premature tube replacements, increased maintenance costs and a reduction in asset life.  

Conversely, underestimating tube temperatures poses significant safety risks. Inaccurate low readings may result in continued operation beyond the metallurgical limits of the tube material, increasing the likelihood of undetected creep damage and, ultimately, catastrophic failure. Accurate measurements ensure that integrity assessments are reliable and that heaters can be run efficiently without compromising safety. This is why choosing a proven tubeskin sensor design, ensuring proper installation and validating data with techniques like IR thermography are vital for long-term operational success.   

This article presented a joint venture to complete uncommon testing of the co-author’s company’s proprietary tubeskin thermocouple sensor technology^b to improve the accuracy and reliability of tubeskin measurements. The outcome of this investigation documented the potential measurement errors associated with inadequate welds on a sensor. A 1/8-in. (3-mm) gap in welding can lead to a temperature overestimation of approximately 30°F (16°C). If the lack of deposition between the weld and the sensor is larger, it is reasonable to assume that the measurement error would also increase. Additionally, the investigation found that when the sensor is shielded, the temperature measurement can be underestimated by approximately 80 °F–150°F (44°C–66°C) lower than the actual temperature, depending on the size of the shielding. Based on this study, training welders to minimize welding mistakes is now part of the normal quality control program.   

DISCLAIMER  

Suncor Energy Inc. and its affiliates (collectively “Suncor”) do not make any express or implied representations or warranties as to the accuracy, timeliness or completeness of these statements, information, data and content contained in this article and any materials or information (written or otherwise) provided in conjunction with this article (collectively, the "information"). This information has been prepared solely for informational purposes only and should not be relied upon. Suncor is not responsible for and is hereby released from any liabilities whatsoever for any errors or omissions in the information and/or arising out of a person’s use of, or reliance on, the information. 

NOTES  

^a WIKA 

^b WIKA’s V-PAD® tubeskin thermocouple 

 

REFERENCES  

1 American Petroleum Institute (API), “API 571: Damage mechanisms affecting fixed equipment in the refining industry,” March 2020. 

2 American Petroleum Institute (API), “API 574: Integrity operating windows, 2nd Ed.,” December 2021. 

3 American Petroleum Institute (API), “API 579: Fitness-for-service,” December 2021.  

4 American Petroleum Institute (API), “API 573: Inspection of fired boilers and heaters, 4th Ed.,” January 2021.  

5 Gonzales, G., “WIKA’s Tefracto-Pad® thermocouple: Accuracy in extreme heat,” online: https://blog.wika.com/us/products/temperature-products/tefracto-pad-tubeskin-sensor-proven-accuracy/  

6 Malhotra, S., A. DeLancey, D. Dutcher, M. Tier and T. Schaefer, “Effect of operating parameters on fired heater tube skin temperature measurement accuracy & the development of new improved tube skin thermocouple,” AFRC 2022 Industrial Combustion Symposium, 2022, online: https://afrc.net/papers/2022/5%20-%20Effect%20of%20Operating%20Parameters%20on%20Fired%20Heater%20Tube%20Skin%20Temperature%20Measurement%20Accuracy%20-%20Malhotra.pdf  

7 American Petroleum Institute (API), API 560, ASME Section V, VIII-1 & IX, ASME B31.3 

8 American Petroleum Institute (API), “API 530: Calculation of heater tube thickness in petroleum refineries, 7th Ed.,” April 2015. 

9 Prueter, P. E., “Damage control: High-temperature creep detection—Parts 1–3,” Inspectioneering, September/October 2022, online: https://inspectioneering.com/journal/2022-10-27/10329/damage-control-high-temperature-creep-detection 

The Authors

Related Articles

From the Archive

Comments