A CASE STUDY ON LOOP SEAL RETURN LEG TO

COMBUSTOR NMEJ FAILURES

BORL BINA – 3X275TPH CFBC BOILERS
Dated 03.02.2012

NON METALLIC EXPANSION JOINTS (NMEJ) is provided for taking up the vibration & thermal
expansion in air & flue gas duct system. The expansion joints are vital components in
thermal power plants. They are installed as flexible connections in air & flue gas duct
system to take up or compensate thermal expansion, vibrations and misalignments.

All the combustor connections have NMEJ to meet multi-directional movements during
operation. While other connections of comparable duty conditions behave as designed, we
faced problem of repeated failure of NMEJ at loop seal return leg to combustor.

Non-metallic expansion joints installed at LOOP SEAL RETURN LEG TO COMBUSTOR (2 nos
per boiler) at BORL are having the problem of damages/ burning in less than 6 months.
Such type of problem was also faced at similar type of CFBC boilers at KLTPS & PT-IBR,
where FBHEs are outside the combustor. However same problem is not faced in the CFBC
boilers, which are having integral FBHEs at SLPP, Barsingsar etc.

ANALYSIS OF FIRST NMEJ FAILURES:

During the visit on June 2011, BHEL Trichy design expert clarified the design improvements
(provision of box bolster instead of ceramic wool blankets) made on the existing
arrangement towards reliable operation of these NMEJs. Further, erection of NMEJs both
RHS & LHS in Unit#1 was done under supervision of M/s KE Burgmann, NMEJ vendor and
every stage, inspection was made. Unit#1 was taken to service on 12 06 2011 and was in
service up to 19 06 2011. During this period of operation, steam load was taken to around
200 TPH as desired by BORL. Further loading was not attempted due to power/ steam
demand limitations. Subsequently, the boiler was restarted on 29 06 2011 and taken on
coal firing on 30 06 2011. In a short while after coal firing LHS combustor to Cyclone NMEJ
was punctured all around and started leaking with gas first, followed by bed ash. (See a
small video clip on NMEJ fuming on 30.06.2011 attached).
Careful analysis of the solid fuel introduction pattern revealed that continuous dumping of
huge quantities of coal with high degree of surface moisture resulted in uncontrollable heat
release resulting in pressure pulsations leading to failure of LHS NMEJ – the weakest link.
While every startup and loading of boilers are to be with careful progressive increase of fuel
feed as per the startup curve, it can be seen from the operating trend, coal feed rate was at
75% speed of Drag Link Chain Feeder continuously without looking at the parameters. From
the pressure pulsations of cyclone bottom, back pass inlet, combustor upper pressure
differential etc., it is very clear that huge quantity of solid fuel has gone into the boiler and
resulted in this kind of total damage of the NMEJ. Following control room snaps provide
details of pressure pulsations and temperature swings at the time of NMEJ failure. Such
swings arising out of uncontrollable heat addition cannot be accommodated by the boiler
system and the weakest section gave way first safe guarding the combustor.
Fuel admission at very high feed rate – @15.7 rpm (maximum speed – 20 rpm)

Combustor Upper Differential Pressure – 10.5 mbar(Normal value <5 mbar)

LHS Stand pipe pressure peaks scaling to approximately 142.8 mbar (Normal value ±5mbar)

Back pass inlet pressure abnormally high @3.12/ 4.26/ 5.21 mbar(Normal value 0.0 mbar)
Damage to NMEJ in other boilers were also reported only during introduction of coal firing
and on the same side of coal feed start. We went ahead improving NMEJ design with a belief
that trained operators managing operation would attach importance and maintain heat
addition rates. However, adoption of this type of uncontrolled fuel input could have been the
only reason for such failures of past.

The instances of damages of NMEJ (Loop seal return leg to combustor) are given below:

BORL – BOILER#1 (Coal firing in November’2010)

Damage Date Replacement Date REMARKS
20.05.2011 24.05.2011 Replaced with other NMEJ fabric lying with BORL & LU
done on 24.05.2011
27.05.2011 05.06.2011 Replaced as per Drg. Rev01 with box bolster
arrangement & LU done on 12.06.2011
30.06.2011 Shut down was not taken, but boiler was run on
minimum load of around 50-60TPH on coal firing.
06.09.2011 Shut down was taken on 16.08.2011 on refractory
problem & NMEJ was replaced with modified design as
per Rev.03 - SS plates, air sealing, box bolster, dust
trap, harmonica bolster etc. LU done on 01.10.2011.
Till date NMEJ is in working condition.

BORL – BOILER#2 (Coal firing in December’2010)

Damage Date Replacement Date REMARKS
20.06.2011 07.08.2011 NMEJ with modified design Rev.03, Light up done on
08.08.2011
21.10.2011 24.10.2011 NMEJ-RHS - outer fabric replaced. Light up done on
24.10.2011
30.12.2011 17.01.2012 Shut down was taken on 03.01.2012, NMEJ-LHS -
complete box bolster, dust trap, harmonica bolster,
Fabric, nuts & bolts etc. replaced. Light up done on
19.01.2011

BORL – BOILER#3 (Coal firing in March,2011)

Damage Date Replacement Date REMARKS
22.06.2011 03.07.2011 Replaced as per Drg. Rev01 with box bolster
arrangement & LU done on 03.07.2011
30.01.2012 NMEJ-LHS started burning. However Boiler operation
continued on coal firing with back pass pressure
maintained –ve and LHS coal feeding reduced.

OTHER SITE EXPERIENCE ON SIMILAR FAILURES:

KLTPS:
• Light up - Oct’08
• COD - Dec’09
• NMEJ Damage – Feb’10
• KLTPS got fabric & bolster replaced through some local agency M/s Urja, but it did
not work and contacted with BHEL Trichy & KE-Burgmann for solution.
• NMEJ replacement with modified design as per Rev.02 (without SS plate & Air sealing
arrangements) - March’11
• NMEJ (modified) damaged – Sept’11
• Outer fabric & bolster replaced through KE-Burgmann.
PTIBR Indonesia:
• COD – Jan’10
• NMEJ Damage – Dec’10
• NMEJ replacement with modified design as per Rev.02 (without SS plate & Air sealing
arrangements) - Jan’11
• Working till date satisfactorily.

NLC 2X250MW:
Unit#1 was commissioned in June’11 and gone up to 180MW load recently. NMEJ is working
satisfactorily. Here solid fuel injection point is given after NMEJ, unlike before NMEJ in case
of BORL Bina & KLTPS.

On frequent failures of loop seal return leg to combustor NMEJ failures,

design expert along with vendor (KE Burgmann) representative and other senior officials of
BHEL visited site in the first week of July’11 and held discussions on addition of pressure
barriers to manage pressure fluctuations during solid fuel introduction. Final
recommendations were discussed with BORL and BHEL proceeded with ordering the material
for immediate implementation in unit#2, which was under shutdown and further phased
implementation of the improvement in other units as well and following were concluded:
- While running the boiler on oil, no damages were found.
- On feeding the coal, smoke was found over the fabric fixing flanges and through the
fabric also. The hot bed material was also coming out from many locations of the
NMEJ.
- It was concluded that hot bed material was touching the expanded harmonica bolster
have punctures and damaged the outer fabric.

Following measures/ improvements were decided to overcome the problem:

• To prevent the coarse hot bed materials entering into the harmonica bolster area:
- One no 10mm thk SS310 ring to be welded on existing gusset plates under floating
ring to have perfect contact with floating ring.
- One no 10mm thk SS310 ring to be placed inside the floating ring to arrest the gap
between floating ring and inner sleeve.
- Box - bolster to be placed inside the floating ring.
- Ceramic wool packing to be done around the floating ring.
- One no 10mm thk SS310 ring to be placed over box bolster to arrest the gap
between inner sleeve and holding plate.
• To prevent fine hot bed materials entering into the harmonica bolster area:
- Arrangement of dust trap seal above box bolster.
• Loose super wool stuffing to be done inside and outside of dust trap seal. Harmonica-
bolsters to be fixed on fabric fixing flange.
• Fabric to be fixed over harmonica-bolster.
ARRANGEMENT OF FLUE GAS PATH FROM COMBUSTOR-CYCLONE-
COMBUSTOR

BACKPASS

CYCLONE CYCLONE
RHS COMBUSTOR -LHS

LOOP SEAL
RETURN LEG TO
COMBUSTOR
SOLID FUEL INJECTION

NMEJs
POINT
 REV.00 (ORIGINAL DRG)

IN DRG. REV.01 – CERAMIC WOOL BLANKET WAS REPLACED WITH

BOX- BOLSTER
 REV.02 installed at KTPS & PT-IBR

1. Fabric, 2. Harmonica Bolster, 3. Loose wool 4. Dust Trap seal. 5. Box Bolster
6. Stud. 7. Clamping bar
 REV.03 installed at BORL BOILER#1&2

Item no 1 Fabric – consists of one layer each of KEB 1400HT, Super-wool – 128kg/cm3 – 12
thk, Super-wool – 128kg/cm3 – 12 thk, Glass fabric – 3 mm thk, PTFE – 0.2, KEB 1300,
KEVLAR, PTFE - 0.2, KEB 1300, SSWM 30X30.
Item no 2 Harmonica Bolster – consists of one layer each of SSWM, HS600 and super wool
insulation.
Item no 4 Dust Trap seal – consists of one layer each of SS HD wire-mesh, High silica fabric
– HS 1250, Super-wool, SS laminated glass fabric, KEB high temp. resistant treated glass
fabric – KEB1400HT.
Item no 5 Box Bolster – consists of one layer each of SS-HD wire-mesh, high silica fabric –
HS1250 and super wool insulation.
NMEJ ERECTION PROCESS FOR DRG REV.03
This following work instruction were prepared to improve the sealing of NMEJ at Loop seal
location against Hot bed material penetration (at BORL Bina site)
1) Refractory under Floating ring to be inspected for its intact condition. If any damage
is found, it is to be repaired properly.
2) One 10mm thk SS310 Plate (Ring-1) is to be welded with the Gussets available
under Floating ring, as per BHEL drawing (Rev.03) This Ring is in 3 segments.
3) Floating Ring “L” Angle to be placed in position as per BHEL drawing.
4) One 10mm thk SS310 Plate (Ring-2) will be kept inside the Floating Ring. This Ring
also is in 3 segments. This Ring-2 width will be nearly similar to Floating Ring Width.
Hence Ring-2 with Floating Ring welding should not be done.
5) Tightly pack the ceramic wool 128 Kg/m3 around the Floating ring (outside only).
6) Place the Box Bolster in position and do the end Joining.
7) Place the 10mm thk SS310 Plate (Ring-3) over box Bolster and tack Weld the
segments at outer edge and middle to form as a single Ring.
8) This assembly (Ring-3 and Box Bolster) to be compressed to the required level by
Hydraulic Jacks.
9) The Cover Plate to be placed in position and to be bolted.
10) Studs for Dust trap seal to be welded by locating with Template.
11) Loose wool stuffing is to be done.
12) Dust trap seal to be fixed in position by using Clamping bar.
13) The upper side (Inlet side) Neck plate need to cut approx 1mtr length for easy
accessing of Dust trap seal to end Join.
14) After end joining of Dust trap seal, the removed Neck plate portion can be welded
with extra care to avoid falling of weld spatters on Dust trap seal.
15) Loose wool (Super wool) to be stuffed around Dust trap seal as shown in drawing.
16) Harmonica bolsters to be placed in position and end joining to be done.
17) The outer fabrics to be fixed in position and to be tighten with Clamping Bar using Gr
8.8 Fasteners.
18) End Joining of Fabric to be done by KEB Supervisor.
19) After reaching the full load running, hot tightening of all Fabric fixing Fasteners is to
be done.
20) The sealing Air connection as shown in the drawing is to be done.
21) All gaps, critical dimensions and steps of erection are to be recorded and protocol is
to be made.
22) All the above works are to be witnessed by BHEL and KEB representatives and the
same are recorded.
23) All fabric fixing fasteners to be placed in position. SS shims to be provided under
Clamping bar ends.
24) No deviation between site and approved drawings unless otherwise agreed by BHEL
and KEB representatives.

COST OF NMEJs (2 SETS as per REV.03) REPLACEMENT FOR ONE BOILER:

Price of 2 sets of modified NMEJ supplied by KE(B) - Rs 10.20 lakhs

SS plates procured by BORL - Rs 8.80 lakhs
Supervision by KE(B) for 15-20 days - Rs 1.50 lakhs
Erection labour cost - Rs 2.00 lakhs
Air purging/ sealing arrangements Rs 1.00 lakhs
Total Rs 24.50 lakhs per boiler

DURATION FOR REPLACEMENT:

15-20 days
 OBSERVATIONS OF MODIFIED NMEJS DURING BOILER
OPERATIONS:

BOILER#2:
- Light up was done on 08.08.2011 after replacement of modified NMEJs.
- RHS NMEJ has started leaking on 20.10.2011 & immediate shut down was taken on
21.10.2011 to avoid further damages.
- The outer entire fabric replaced and light up was done on 24.10.2011.
- LHS NMEJ started leaking on 31.12.2011, but boiler was continued till 03.01.2012 as
other 2 boilers were also in shut down condition.
- The entire fabric, harmonica bolster, box bolster & dust seal and nut & bolts were
replaced and boiler light up was done on 19.01.2012.

BOILER#1:
- Light up was done on 01.10.2011 after replacement of modified NMEJs.
- Till date, it is working satisfactorily except some minor outer most SS wire-mesh
damage due to falling of coal dust, which was rectified in the last week of Jan’12.

All possible modification, which can be thought off after discussions between KE(B)
designers, BHEL Trichy designers & BHEL Site/ BORL engineers for 2 days in the month
of July’11, were carried out in the loop seal return leg to combustor NMEJs, but NMEJ
failed in less than 3 months of boiler operation.

On taking up the issue of NMEJ failures, vendor M/s KE Burgmann washed off their
hands and instead offered another design (REV.04) to BHEL Trichy:

1. Outer fabric, 2. Outer curtain, 3. Inner curtain, 4. SS shims, 5. Fabric fixing

fasteners, 6. Box Bolster, 7. Canopy

Air purging arrangements are made from top, unlike from bottom in the REV.03 and
harmonica bolster has been removed.
CHANGES MADE DURING RECENT RECTIFICATION OF MODIFIED NMEJ (LHS)
Air purging/ sealing arrangements made in Boiler#1&2 as per drawing Rev.03
were disconnected on the recommendation of BHEL Trichy after the failure of NMEJ
(LHS) and seeing the fine dust accumulation, which was more than usually seen without
air connection. M/s KE(B) did not recommend for the air purging arrangement from the
top on the existing design NMEJs as it is possible only in the latest proposed modification
Rev.04.

Removal of entire wire mesh from box bolster from shell side as undulations of
Box bolster at shell side were observed and based on the feedback from KLTPS site.

PHOTGRAPHS OF NMEJS

Boiler#1 NMEJ burnt out on 30.06.2011

Boiler#2 NMEJ (LHS) – Red hot bed material can

be seen inside fabric on 02.01.2012
BOILER#2 NMEJ (LHS) PHOTOS AFTER DAMAGE ON 30.12.2011

Damage of outer fabric

Damage of Harmonica Bolster

Fine Dust accumulation on & around dust trap seal

NMEJ PHOTOS DURING ERECTION AS PER DRG REV.03

Box bolster during NMEJ erection for boiler#1

Dust trap seal erection

 Harmonica bolster erection

Final outer fabric erection

VIDEO CLIPS OF NMEJ DAMAGES:

1. Boiler#1 NMEJ, Fumes coming out on 30.06.2011.
2. Boiler#3 NMEJ (LHS) on 31.01.2012
3. Boiler#3 NMEJ (LHS) on 03.02.2012