Vertical Tube, Variable Pressure Furnace for Supercritical Steam Boilers

D.K. McDonald The Babcock & Wilcox Company Barberton, Ohio, U.S.A. S. S. Kim U.S. Department of Energy National Energy Technology Laboratory Pittsburgh, Pennsylvania, U.S.A.

Presented to: POWER-GEN International 2001 December 11-13, 2001 Las Vegas, Nevada, U.S.A.

BR-1723

Abstract
The U. S. Department of Energy (DOE) is developing highly efficient, low-emission, and low-cost power generation technology with industry participation for future power plants. As part of the DOE advanced combustion program, The Babcock & Wilcox Company/McDermott Technology Inc. (B&W/MTI) will be testing a system that emulates a vertical tube, variable pressure furnace designed for supercritical boilers. Two types of supercritical boiler furnace designs are in use today, both requiring high mass flow per tube for cooling. One operates at a constant fluid pressure in the furnace and utilizes vertical tubes, and the other allows the fluid pressure to vary with changing loads and utilizes a spiral furnace design. The constant pressure furnace design has been used primarily in the U.S. while the variable pressure design has been dominant in Europe. B&W/MTI has completed the design of a system to test the low mass flow, vertical tube, variable pressure furnace design at Allegheny Energy Supply Companys Hatfields Ferry Unit #3, a 575 MWe B&W supercritical boiler. Test panels, approximately 100 feet long, will be installed on the front and sidewalls of the boiler that provide differing heat flux patterns. The testing will demonstrate the flow characteristics and dynamic stability of the optimized multi-lead ribbed tube that is key to this furnace design, and validate the models for theoretical analysis and design. The low mass flow vertical tube furnace design will allow variable pressure operation, on/off cycling, and rapid load changing, that are the same operating advantages seen in a spiral design. This vertical tube design, however, will have additional benefits over the spiral design that include lower fabrication cost, lower

construction cost, 45% lower furnace pressure drop (i.e., less feed pump power needed), increased operating margin against tube overheating, and less maintenance cost. Startup and testing is planned for the spring-summer of 2002. In this paper, design features of the optimized multi-lead ribbed tube and the test plan will be presented.

Introduction
A variety of political and financing pressures stemming from global concern about emissions, including greenhouse gases, are elevating interest in improving plant efficiency. Although the recent U.S. domestic trend has been toward natural gas fired combined cycle (CC) peaking plants, the increasing cost of natural gas, the need for fuel diversity, the abundance of inexpensive coal and the increasing price and schedule for combined cycle plants have revived interest in coal and new coal-fired plants in the U.S.. As a result, boiler manufacturers are receiving an increasing number of specifications for boilers, including oncethrough, supercritical boilers with increasing steam temperatures and state-of-the-art emissions controls. At the same time, low power margins and regional shortages are imposing tremendous pressure on suppliers to minimize costs and shorten schedules. In order for advanced supercritical plants to provide the desired efficiency and emissions and be attractive in the U.S. and abroad they must be both demonstrated and low cost. Therefore, investment in development of advanced supercritical plants must be directed toward both performance and cost reduction. Through a program sponsored by the U.S. Department of Energy (DOE),

Babcock & Wilcox

ing installation, testing of the system will commence consisting of operation over the load range to replicate the furnace heat flux, pressure, flow, and temperature conditions expected in a vertical tube, variable pressure, supercritical boiler that employs the low mass flow furnace design.

Background
In order to improve coal-fired power plant efficiency leading to a proportional reduction in coal consumption and carbon dioxide emissions (Figure 2), it is widely accepted that the domestic power industry must move from subcritical to supercritical steam cycles. The supercritical design not only improves efficiency by increasing the working fluid pressure but it allows superheating of the steam to higher temperatures which provides significant further efficiency improvement. For many years the most popular boiler design in the U.S. has been the subcritical drum boiler. This technology is low cost and well proven but does not have the potential for efficiency improvement inherent in supercritical cycles. Two types of supercritical boiler designs are currently in use, those that operate with a constant pressure in the furnace tubes and those that vary pressure with load. The latter is the popular design for supercritical boilers today because it is not only more efficient at lower loads, but in combination with a circulation pump, it can also be cycled on and off much more rapidly. These features permit the owner to more effectively follow system demand. The constant pressure furnace has been used primarily in the U.S. while the variable pressure furnace design has been dominant in Europe (see Figure 3). The constant pressure furnace utilizes tubes that are vertically oriented while the variable pressure furnace design wraps the tubes around the furnace in a spiral fashion. Since the late 1950s B&W has supplied 90 supercritical boilers representing 60,000 MW (42,000 coal-fired) operating with constant pressure furnaces ranging in size from 343 MW to 1300 MW. Boilers with constant pressure furnaces can accommodate variable pressure operation by utilizing a division valve between the furnace and superheater to control pressure to the turbine. B&W has also supplied both sub and supercritical variable pressure once-through boilers with spiral furnaces. In a rectangular furnace, the heat flux imposed upon the furnace walls is not evenly distributed as shown in Figure 4. To accommodate the differences in heat flux around the periphery of the furnace and the upsets inherent in the combustion process,

Figure 1

Hatfields Ferry Unit 3.

Babcock & Wilcox (B&W) is conducting testing to demonstrate a new low mass flow furnace design capable of variable pressure operation that will reduce the cost and improve the efficiency of once-through boilers. Demonstrating the advanced circulation design of the vertical tube variable pressure furnace by testing panels represented by single optimized multi-lead ribbed tubes (OMLR) in a real boiler furnace will enable use of the technology in the next generation of supercritical boilers. To verify the design, a demonstration system has been designed for installation at Allegheny Energy Supplys supercritical 575 MW Hatfields Ferry Unit 3 (Figure 1) in the spring of 2002. B&W is providing project coordination, engineering for design, fabrication and installation of the equipment as well as conducting the testing including collection, analysis and evaluation of the data and final reporting. Contract management is provided by B&Ws affiliate, McDermott Technology, Inc. in Alliance, Ohio. The new OMLR tube has shown the characteristics needed to achieve natural circulation at low mass flows in laboratory tests. In once-through boilers the mass flow is normally too high to allow the flow to naturally redistribute to the tubes with higher heat input to protect them from overheating. The objective of this program is to demonstrate the natural circulation characteristics and the dynamic stability of parallel flow circuits composed of OMLR tubes at low mass flows in the field at full scale. Follow-

&2 (PLVVLRQV W\U

            



$VVXPSWLRQV  0:  %WXOE ++9  &DUERQ LQ &RDO  &DSDFLW\ )DFWRU  &RPEXVWLRQ (IILFLHQF\
  

+HDW 5DWH (Btu/kWh) Heat Rate

Figure 2 CO 2 reduction versus plant heat rate.

Babcock & Wilcox

3rd Pass

2nd Pass

Mix

1st and 2nd Pass

1st 1st Pass Pass

B&W UP Boiler (Constant Pressure Furnace)

Figure 3 Once-through boiler furnace arrangements.

SpiralBoiler (Variable Pressure Furnace)

sufficient fluid flow must be provided to each tube to maintain tube temperatures within acceptable limits over the load range. To provide sufficient flow per tube, constant pressure furnaces, such as B&Ws Universal Pressure (UP) design, employ vertically oriented tubes appropriately sized and arranged in multiple passes in the lower furnace where the burners are located and the heat input is high. By passing the flow twice through the lower furnace periphery (two passes), the mass flow per tube can be kept high enough to ensure sufficient cooling. In addition, the fluid is mixed between passes to reduce the upset fluid temperature. The spiral design, on the other hand, utilizes fewer tubes to obtain the desired flow per tube by wrapping them around the furnace to create the enclosure. This also has the benefit of passing all tubes through all heat zones to maintain a nearly even fluid temperature at the outlet of the lower portion of the furnace. Figure 5 shows schematic furnace arrangements for the high mass flow constant pressure vertical tube and spiral variable pressure designs as well as the new advanced low mass flow vertical tube variable pressure design. Because the tubes are wrapped around the furnace to form the enclosure, fabrication and erection are considerably more complicated and costly. Figure 6 shows the complexity of the tube arrangement forming the spiral and the additional vertical bars
@IUC6GQ`Q6S6GG@GTC@6U

287/(7 (17+$/3<

+($7 )/8; ',675,%87,21 )521 6,' FRONT SIDE

Figure 4

5($ REAR

SIDE 6,'

Once-through boiler heat distribution.

required for wall support. Because of its relative simplicity and ability to self-support, the vertical tube furnace design is significantly less costly and easier to fabricate and construct than the spiral variable pressure furnace. Over absorption of tubes receiving more heat than the average is an inherent capability of a natural circulation drum boiler because the tube mass flow increases with more heat input to redistribute the water flow from tubes with less heat input. A high mass flow once-through boiler will have a forced circulation characteristic wherein the flow decreases with increasing heat input. As shown in Figure 7, with natural circulation tubes receiving more heat than the average tube naturally draw more flow, which increases cooling and protects the tube from overheating. However, in a once-through forced circulation furnace with high mass flow the tube receiving more heat than the average tube receives less flow, which can result in further increasing temperatures, potentially leading to failures. Thus the mass flow per tube must start very high to ensure adequate remaining flow after the heat upsets. Designs with medium mass flow, though attempted in oncethrough forced circulation boilers with moderate success, exhibit worse consequences than the high mass flow designs. When the mass flow is degraded during load reduction in a tube receiving more heat than the average, the remaining flow will have less margin to provide acceptable cooling. Medium mass flow designs can experience heat upsets and/or flow excursions that result in flows in individual tubes that are lower than the low mass flow design. A design with low mass flow, in the same range as used in subcritical drum type boilers, will be capable of natural circulation which will increase rather than decrease the mass flow in tubes receiving more heat than the average thus protecting the tube from overheating. The ideal furnace design for a once-through boiler would have vertical tubes and capability to operate with variable pressure operation over the load range while exhibiting natural circulation characteristics, thus protecting the tubes from overheating. These

Babcock & Wilcox

UG 3DVV 0L[ +HDGHU 0L[ +HDGHUV VW 3DVV 'RZQFRPHU :DWHU ,QOHW 7UDGLWLRQDO 7ZR 3DVV /RZHU )XUQDFH :DOO :DWHU ,QOHW

UG 3DVV

UG 3DVV

6SLUDO /RZHU )XUQDFH :DOO

0L[ +HDGHU $GYDQFHG 20/5 /RZHU )XUQDFH :DOO 7XEHV :DWHU ,QOHW

83 %RLOHU /RZHU )XUQDFH 'HVLJQ KLJK PDVV IOX[

6SLUDO /RZHU )XUQDFH 'HVLJQ KLJK PDVV IOX[

$GYDQFHG /RZHU )XUQDFH 'HVLJQ ORZ PDVV IOX[

characteristics should be achieved with as low a flow per tube as practical to obtain a low friction resistance to promote natural circulation, minimize furnace pressure loss, and reduce the pump power required. Such a design represents a significant advancement for tube protection in once-through boiler circulation technology. Achievement of sufficient natural circulation characteristic in a once-through furnace to maintain safe temperatures in all tubes and allow for variations of heat input, while operating with variable pressure and minimal average mass flow, has not been adequately demonstrated to produce the desired effects. Until recently, the components and design criteria to provide a successful furnace design which permits a low mass flow per tube, operates with variable pressure over the load range, is capable of onoff cycling, and has all of the structural, fabrication and installation advantages of a furnace with vertically oriented tubes has not been available.

Description of the System

The success of the low mass flow vertical tube variable pressure design depends on the capability of the tube internal geometry to promote cooling of the tube when exposed to high heat fluxes. Two rib designs, single and multi-lead, have been applied in the past. The single lead design promotes significant turbulence at the wall and is excellent in preventing departure from

nucleate boiling (DNB) but it also produces a high pressure drop due to friction and must be produced by machining, which is expensive. The conventional multi-lead (MLR) tube provides a much lower friction resistance (and pressure loss) and can be extruded but does not produce as much turbulence at the wall as the OMLR tube, which is needed to prevent DNB at flows low enough to achieve natural circulation characteristics in a oncethrough boiler. The primary application of MLR tubes is to prevent DNB in high heat flux zones of boilers to reduce the maximum average mass flow per tube required for the furnace design. To provide sufficient cooling at a low mass flow per tube, Siemens-KWU built on previous work by B&W that provided known performance characteristics for internally ribbed tubes to develop their patented optimized internally ribbed tube (OMLR) design. This tube has displayed natural circulation characteristics and excellent cooling at low flow. Though it can be fabricated using normal extrusion methods, the rib design is unique in that it employs a greater rib height and a flatter lead angle which creates turbulence and mixing near the tube wall to promote cooling (see Figure 8). B&W is now taking this concept to the next step of more complete testing of the tubes characteristics at full scale in a real boiler furnace. Successful application of the vertical tube variable pressure furnace design will provide all of the operational benefits of the currently popular spiral boiler design while significantly reducing the cost and construction time for the furnace and providing some reduction in pressure drop and an increase in plant efficiency.

Support Bars

Figure 6

Spiral lower furnace.

Figure 7 Once-through versus natural circulation characteristics.

Babcock & Wilcox

Increased LQFUHDVH Rib Height

G LE

Increased UHGXFHG Rib Height

O G

Figure 8

6LQJOH /HDG Ribbed Tube 5LEEHG 7XEH Ribbed (rifled) tube designs.

Single Lead

Tube (MLR) 0XOWL /HDG 5LEEHG 7XEH 0/5

Multi Lead Ribbed

Optimized Multi Lead Ribbed Tube (OMLR) 2SWLPL]HG 0XOWL

/HDG 5LEEHG 7XEH 20/5

To test this concept, a system of single tubes representing furnace panels has been designed for installation inside the existing membrane wall of Allegheny Energy Supplys Hatfields Ferry Unit 3 in the spring of 2002. The design of this system has been completed and was reviewed with Allegheny Energy Supply prior to the fabrication and procurement phase, which is now in progress. Appropriate provisions have been made for support, thermal expansion, and routing of the instrumentation. The test tubes or panels will be within the front and sidewalls at locations that provide the desired length and heat flux pattern. The final positions of the tubes have been selected to avoid interference with burners, access doors, sootblowers, tempering ports, and control and monitoring instrumentation used by the operators. As Figure 9 shows, the test tubes begin and end at headers that are individually connected to the source and place of discharge of the incoming and exiting fluid. A safety valve is provided on the outlet bottle between the isolation valves, as re-

quired by code, and vented to the outside of the building. The basic flow circuit arrangement consists of two possible flow paths, one for test operation and the second for non-test operation of the test tubes. During testing the main inlet water flow will be taken from the first pass downcomer and mixed with lower temperature water from the discharge of the boiler feed pump. For most of the time the test tubes are installed, the panels will be operated in the same manner as the second pass circuits in the existing furnace wall (constant pressure). During these periods, the main water flow for the test system will be provided from the first pass downcomer and will flow to the inlet bottle for distribution to the test tubes representing furnace panels. The outlet flow from the tubes will be routed to the second to third pass mix bottle. Allegheny Energy Supplys operators will control the units operation at all times and B&W will monitor conditions in the test system closely during testing to achieve proper conditions.

Furnace Arch Side Wall Te s t Tubes Front Wall Te s t Tubes

Figure 9

Test system arrangement.

Babcock & Wilcox

  ,QOHW 3UHVVXUH SVLD             3HUFHQW /RDG
Figure 10 Test system pressure ramp relative to the host furnace.

)XUQDFH 7HVW 3DQHOV







During non-test operation, the test system will have no impact on unit efficiency or operation. During testing, the lower pressure steam exiting the test system will not be available for power production but heat will be recovered in the existing steam coil air heater to minimize losses. This amount of flow will be less then 0.4% of steam flow over the 20% to 100% load range of the test. Therefore, when testing, the impact on the unit will be negligible. The testing program is designed to verify the aspects of commercial operation including the flow characteristics of the tubes, variable pressure load increase and decrease capability and rate as well as dynamic characteristics during load changing and stability at minimum load. Figure 10 compares the operating pressure in the test system versus the operating pressure in the host boiler furnace tubes.

Summary of Progress
Work on the program began in August 2000 with initiation of engineering of the circulation system including the thermal-hydraulic and heat transfer analyses of the test panels. A plant visit was made in September to begin the information exchange with the host that was necessary for arrangement of the system. The plant management was provided with an overview of the plan and schedule and communication channels were established. Analysis of the heat transfer soon determined that a single tube nested between the existing boiler wall tubes represents an OMLR tube in a conventional membraned wall configuration. Initially the system was to be installed during a 4 week long Unit 1 outage scheduled for the spring of 2001. However, as a result of evaluation of their outage plans for all three units in October 2000, the Unit 1 outage was moved to the spring of 2002. At that time it was decided to shift the project from Unit 1 to its identical sister Unit 2 that had a 3 week fall 2001 outage planned. Further consideration of the design changes due to differences between Unit 1 and 2, the shorter outage, as well as the schedule and timing of funding requirements resulted in a decision by B&W in December to return to the Unit 1 outage in the spring of 2002. On December 7, 2000 a Project Kick-off Meeting was held at the DOE National Energy Technology Laboratory (NETL) office in Pittsburgh, Pennsylvania. In addition to a description of the system and test plans, the impact of the outage change on the

system design and project schedule was explained and discussed. Meanwhile work continued on the piping routing and system arrangement. Discussions about the control strategy were initiated and work began to determine the hardware and software necessary to provide the controls required for the testing and longer-term operation of the test sections. In February 2001, Allegheny Energy Supply requested consideration of installing the test system in Unit 3 rather than Unit 1 with no change in schedule. They were considering further optimization of their outage schedules by switching the spring outage to Unit 3 instead of the fall and moving the Unit 1 outage to the fall. They also indicated the outage would be 5 instead of 4 weeks long, which would benefit the program. A site visit was made to determine the impact on the system arrangement to suit Unit 3 and it was not found to be significant. As a result the outage schedule was revised and changes to the drawings were made to accommodate Unit 3. As a condition of hosting the project by Allegheny Energy Supply that became a commitment to DOE as well, a System Review was required prior to release for purchasing of the hardware. After an internal review, a meeting was held at the plant on April 27, 2001 and as a result B&W was approved to proceed to procure the equipment. Specifications were completed for the key components and vendors were contacted to resolve any technical issues and determine cost and delivery schedules.

Plans for the Future

During the fall of 2001 material procurement is in progress with all materials expected to be on site by the middle of February 2002. During November and December 2001 plans for installation will be finalized with Allegheny Energy Supply and the outage contractor. Startup is expected to begin in April 2002 followed immediately by testing. Following testing the system will be left in the normal mode of constant pressure operation until the next outage when all equipment will be removed and the unit restored to its prior condition.

Summary
Currently two furnace designs exist for once-through forced circulation boilers, the constant pressure vertical tube design and the variable pressure spiral wound design. Both employ a high

Babcock & Wilcox

mass flow per tube to maintain tube temperatures within acceptable limits in areas of high heat flux. The less costly vertical tube arrangement must operate at constant pressure, utilizes multiple passes and mixes, does not permit daily on/off cycling and exhibits a slightly higher pressure drop than the spiral furnace design at lower loads. On the other hand, the spiral design is costly to fabricate, erect and maintain. The ideal furnace design for oncethrough boilers would allow the operating flexibility of the variable pressure furnace including on/off cycling and minimal low

load pressure drop, while using a conventional vertical tube arrangement to minimize cost. To accomplish this a tube with optimized characteristics that allows a low mass flow per tube and promotes natural circulation characteristics to enhance tube protection in areas of high heat flux is needed. B&W, with support from the Department of Energy, is testing a tube believed to have the required properties. Successful testing and implementation of this design will take furnace circulation technology to a new level of sophistication.

Copyright 2001 by The Babcock & Wilcox Company, a McDermott company. All rights reserved. No part of this work may be published, translated or reproduced in any form or by any means, or incorporated into any information retrieval system, without the written permission of the copyright holder. Permission requests should be addressed to: Market Communications, The Babcock & Wilcox Company, P.O. Box 351, Barberton, Ohio, U.S.A. 44203-0351. Disclaimer Although the information presented in this work is believed to be reliable, this work is published with the understanding that The Babcock & Wilcox Company and the authors are supplying general information and are not attempting to render or provide engineering or professional services. Neither The Babcock & Wilcox Company nor any of its employees make any warranty, guarantee, or representation, whether expressed or implied, with respect to the accuracy, completeness or usefulness of any information, product, process or apparatus discussed in this work; and neither The Babcock & Wilcox Company nor any of its employees shall be liable for any losses or damages with respect to or resulting from the use of, or the inability to use, any information, product, process or apparatus discussed in this work.

Babcock & Wilcox