Walters Forensic Logo  Publications Contact Us Home 
 

BOILER TUBE FAILURES

Two Scotch Marine type steel boilers made up the power plant for a hot water radiant heating system in a high rise building.

Repeated failures occurred in the boilers over a period of about one year. The failures typically occurred at the start of the second pass where the fire tubes met a tube sheet. Cracks developed which allowed water to escape from the water side into the fire side of the boiler. In each case, magnetic flux tests in addition to dye penetrate tests were conducted to locate the cracks to be repaired. All repairs were conducted under the supervision of the Department of Labour, Mechanical and Engineering Division and were performed in accordance with requirements of the Canadian Standards Association (CSA) B51, American Society of Mechanical Engineers (ASME), and National Board Inspection Codes.

Cracks were repaired in one of the hot water boilers on three separate occasions. The cracks had developed at the back of the boiler where the second pass tubes are welded to the tube sheet. Cracks had developed through the tube, through the weld, and into the tube sheet in each case. Similar cracks were also repaired in the other hot water boiler on one occasion.

The nature of these cracks were indicative of thermal shock.

Thermal fatigue may result in either low cycle or high cycle failure. Low cycle thermal fatigue usually is associated with large plastic stains and most often is caused by large changes in temperature or large differences in thermal expansion between two structural members. If a change in temperature is especially severe, and if it occurs rapidly, it is called thermal shock. Failures due to thermal shock generally occur at ten thermal cycles or less, sometimes in only one thermal cycle when the material is sufficiently brittle.

Pressure and thermal cycling during startup and shutdown are the most common causes of low cycle fatigue failure. Design factors that concentrate strain, influence low cycle fatigue as well as high cycle fatigue. However, the effect is not quite the same. When surface stresses are high enough to be conductive to low cycle fatigue, the presence of a severe stress raiser often will cause a complete fracture to occur on the first load cycle. Mild stress raisers usually do not lead to fracture on the first cycle, but only induce localized yielding. The resultant plastic flow may distribute the stress, but usually not enough to avoid further yielding on successive cycles.

The mechanics of thermal shock typically result in crack initiation on the water side of the tube sheet. In the reported cases, when repairs were effected, either magnetic particle inspection or liquid dye penetrant inspection or liquid dye penetrant inspection was used to detect and isolate the cracks to be repaired. The principal limitation of the liquid penetrant method of inspection is that discontinuities (cracks) must be open to the surface. Similarly, the limitation of magnetic particle inspection is that it is not completely reliable for locating discontinuities that lie entirely below the surface. Consequently, only the cracks which had completely penetrated the tube sheet allowing water to escape to the fire side of the boiler would be detected by these methods. It is likely that additional cracks which initiated in the tube sheet and were not detected, subsequently grew while the boiler was in service and penetrated the tube sheet at a later date, thus necessitating additional repairs.

An investigation revealed that the boilers were operated at temperatures below the manufacturers and American Society for Heating, Refrigerating Air Conditioning Engineers (ASHRAE) minimum recommended levels. It was concluded that operation of the boilers at reduced temperatures over an extended period likely resulted in the thermal cracking experienced.

 
The information contained in this web site is intended for marketing purposes only. It is not all-inclusive, and does not fully describe the many and varied services that the company provides, nor does it completely describe the education, training, skills, or expertise of our staff.

 
 
 

Walters Forensic Engineering | 277 Wellington Street West, Suite 800 | Toronto, ON M5V 3H2
Information contact: engineering@waltersforensic.com