HOME > flue gas expanders

RTI Power Generation Division

THE POWER GENERATION DIVISION OF REFINERY TECHNOLOGY, INC.

RTI EXPANDER PIPING EXPERIENCE

Refinery Technology, Inc. (RTI) has been a part of 31 different expander installations, around the world. On many of these projects, RTI was asked to provide engineering services that would redesign the piping system to avoid future catastrophic failures. RTI was contracted by Phillips 66 Company to redesign an 84"Ø expander inlet line (the largest expander inlet line in the world). The system had been in service eight years and over that time due to various expansion joint failures and piping failures, the FCC Unit had to be shut down four times to make unscheduled emergency repairs, which were very costly in lost production.

Since RTI completed the redesign in 1989, they have had no emergency unscheduled shutdowns to repair the expander piping. All of the expander piping systems that we have been contracted to redesign were originally engineered by major E&C companies and were thought to be appropriate designs; but time proved otherwise.

The high temperature expander piping system is the one piping system that can destroy production profits that a refinery may think that they will make during the years of operation, from one turnaround to another. The losses are extremely high to a refinery that must shutdown and make major repairs to the piping, the expansion joints or the expander itself. This piping system should be treated like open heart surgery, where one seeks the most skilled professional to perform the surgery, likewise, the refining company should find and hire the most experienced and the most knowledgeable company in the world to contribute their design expertise and understanding gained over many years of experience to work out the design details for the improvements of this extremely critical high temperature piping system.

The major design considerations relating to the long term safety, reliability and maintainability of this piping system, of course involve the piping configuration; the design, application and manufacture of the expansion joints; the placement, design and supplier of the spring supports; the placement and design of the pipe guides; and the way in which the pipe stress analysis is modeled, analyzed and interpreted. Make no mistake, any one of the above subjects not engineered correctly will lead to eventual failure.

RTI has specialized for 25 years in the above services, and we have seen and reviewed many expander piping system details that looked good on paper, but RTI was contracted to correct the failed piping or components and make things right to avoid future operating problems.We would like an opportunity to discuss both expander and piping issues with you; so please contact us at your earliest convenience.

AN IMPORTANT MESSAGE ABOUT YOUR EXISTING EXPANDER
If you have ever experienced a blade rub or blade fouling:

We, at Refinery Technology, Inc. (RTI), have realized by actual field test data that the expander flange allowables (inlet and exhaust) that are submitted with each new or rebuilt expander have never been confirmed by actual load tests by the manufacturer. Without the confirmed load test data to validate the allowable flange loads, you, the owner, have no way of knowing how accurate the information from the manufacturer really is. If the expander has experienced a blade rub, the manufacturer may have told you that the piping loads on the expander were too large. Another cause of blade rub could be that the expander may not be strong enough to withstand the manufacturer’s allowable loads at full operating temperature without distorting the case. RTI has developed an accurate and safe expander load test to document the actual change in blade tip clearance. These load tests need to be scheduled during your next regular FCCU shutdown or before, if the unit comes down for 7 or 8 days for any reason they could be accomplished. RTI can also perform a piping load test to determine the true piping loads on the expander. Then, by comparing the results of the expander load test with the true piping loads we can quickly determine if the two can be bolted together to be a reliable system. If you are interested in clearing up the questions relating to your expander problems then e-mail RTI at sales@refinerytech.com or you may call us or fax us at Tel: 281-580-3773, Fax: 281-580-3633. We will respond back to you immediately regarding these important services.

RTI GENERAL INFORMATION REGARDING ALLOWABLE FLANGE LOAD VERIFICATION FOR FLUE GAS EXPANDERS

If you own and operate a flue gas expander or you are in the process of buying a new expander there is something you should strongly consider:

As an owner, or owner to be, you should contact the expander manufacturer and confirm that the new/existing expander is or will be designed to operate at full design temperature for the submitted allowable flange forces and moments. Once it is confirmed that the expander is designed to meet the allowables at the design temperature, then the next documented question should be, “How much deflection would be expected between the shroud and the blade tips if the allowable cold loads were applied to the expander flanges?” The number from the manufacturer is required for record and this is the number that will be used to judge the failure or success of the actual flange load test on the expander prior to shipment, or in the field on an existing unit.

In the expander manufacturer’s shop, or in the field, a cold flange load test must be made. For this test, loads will be applied to the inlet and the exhaust flanges of the expander to produce forces and moments on the flanges that replicate the cold allowable values.

While these forces and moments are being exerted on the flanges the clearances between the inside of the shroud and the tip of the rotor blades will be monitored. The first recorded observation to make is to find the load that is required to deflect the shroud 0.004" or 0.1mm a typical amount given by manufacturers. The next recorded observation to make is what deflection is measured at the shroud to blade tip when the full cold allowable load is applied.

All of the expander manufacturers state that their equipment is built to operate reliably at the design temperature for their submitted allowable loads, but none publish what their calculated shroud deflections are expected to be if tested. What the cold flange load test will establish for the purchaser of the equipment is that the equipment either passes or fails the shroud deflections as submitted by the vendor for the allowable loads. This is the only confirmation that the new owner will get that will verify that the expander is truly designed to meet the allowable loads, as stated by the manufacturer.

The shop or field flange load tests on the expander will be conducted on a cold (ambient) machine, but the submitted vendor allowable loads are to be used for an operating (hot) machine at operating temperatures (1350°F / 732°C or higher), when the casing material is at its weakest strength. So the loads used for the cold flange load test must be adjusted upward by the ratio of the cold to hot modules of elasticity (approximately 1.4).

EXPANDER INLET LINE PRESSURIZED PULL TEST

Any refinery that has an operating hot gas expander should strongly consider having an “Inlet Line Pressurized Pull Test” completed during their next scheduled turnaround. This test will take up to three days to complete.

Why is this pull test so very important to the refinery? Because this test is an actual representation of the real piping forces that exist on the expander inlet casing flange as the piping heats up to operating temperature during the start-up mode or for any thermal increase during an excursion condition.

The general configuration of an expander inlet piping system is as follows: It comes off the third stage separator and runs horizontal either on the centerline of the expander or perpendicular to the centerline over to a point, then it drops down vertically through a tied double hinge expansion joint or a tied double gimbal expansion joint. Then it turns horizontal and runs straight to the expander inlet casing flange. So that it’s not forgotten, there is always another single hinges expansion joint that must work in conjunction with the previously mentioned expansion joint (s).

By performing this pull test we are able to measure the actual force required to move the expansion joints under pressure, simulating their thermal deflections as though the stainless steel piping system was heating up to the operating temperature.

The major difference with the pressurized pull test versus non-pressurized is that the expansion joint bellows with internal pressure now produces pressure thrust forces that act on the expansion joint hinge or gimbal pins and the pressure thrust force can be in the range of 50 to 130 thousand pounds depending on the mean diameter of the bellows and the operating line pressure. If the hinge or gimbal pins are not designed properly to minimize the pin friction the expansion joint will offer resistance to the thermal movements, which will generate large axial forces on the expander inlet casing through the entire heat up cycle during start-up.

These forces that can be measured and documented as the result of the pressurized pull test can then be added to a detailed pipe stress analysis for a start-up case study to determine a much more realistic and accurate expander nozzle load, which should be presented to the manufacturer for approval.

The computer generated loads will vary greatly from the actual pressurized piping pull test loads. Case in point: RTI was asked to audit a completed computer analysis for an existing expander inlet line where the expander had experienced some problems thought to be piping related. The computer analysis model was done correctly with the available data input information. The theoretical analysis revealed an axial load force of approximately 2600 lbs. RTI recommended that a pressurized piping pull test be conducted. The pressurized pull test revealed a real axial force load of 6500 lbs. Now, with the problem revealed the course of action was defined and completed by RTI.