Service Experience 2008, MAN B&W Engines

ME/ME-C Engine Series

Contents: Introduction .............................................................................. 3

Update on Service Experience ................................................ 4

– Hydraulic cylinder unit (HCU) ................................................... 4
– Multi purpose controller (MPC) ................................................ 10
– Hydraulic power supply (HPS) ................................................. 12
– Servo oil system ...................................................................... 12
Conclusion ................................................................................ 13

MAN Diesel A/S • Copenhagen, Denmark

Service Experience 2008, MAN B&W Engines
ME/ME-C Engine Series

Introduction 1000
-30% flow area +10mm shims , 13ms step-profile, -20bar Pcomp
900
The number of electronically controlled
engines in service continues to grow 800 -30% flow area,
and, at the time of writing, more than +10mm shims
500 engines are on order or in service. 700 -30% flow area
+10mm shims,
During 2006 and 2007, “ME optimi- 600 13ms step-profile,
bar abs

sation” by variation in parameters -10bar Pcomp

was carried out on our test engine, 500
4T50ME-X, on 12K98ME, 6S60ME-C
400
and, not least, on the 6S40ME-B pro-
totype engine. Parameters which were 300
varied are:
200
Reference
- Injection timing 100

0
- Exhaust valve timing 170 180 190 200 210
deg a BDC
- Compression volume configuration
Fig. 1: Injection profiling
- Fuel valve flow area

- Injection profile (step-profile)

3.0
An example of injection profiling (“rate- ME --- SFOC
shaping”) can be seen in Fig. 1. 2.0
Fig. 2 shows an example of the result of Reference
C

parameter variation resulting in an im-

NO x in g/kWh

1.0
FO
x /S

proved NOx /SFOC relationship.

ME --- NO x
O
-N

0.0
--

At the end of 2007, the first S40ME-B -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0
E
M

engine was prototype-tested at STX in -1.0

Korea, Fig. 3. These tests mark the be-
Relative

ginning of an era where the full potential -2.0

of the electronic fuel injection with “rate
shaping” (or “injection profiling”) is uti- -3.0
lised on production engines giving a
very attractive NOx/SFOC relationship. -4.0
Relative SFOC in g/kWh
At the beginning of January 2008,
the first four LNG carriers with Fig. 2: 6S60ME-C parameter test 75% load
2X6S70ME-C engines (Fig. 4) were
in service. During 2008, this number
will increase to 20 vessels.

3
 Update on Service
Experience, ME/ME-C
Engine Series
This paper focuses on recent service
experience for ME/ME-C engine series.
At the end of 2007, 130 ME/ME-C en-
gines were in service. The reporting will
be divided into the various sub-systems
of the ME/ME-C engines. These are the
Hydraulic Cylinder Unit (HCU), the Multi
Purpose Controller (MPC), the Hydrau-
lic Power Supply (HPS) and the Servo
Oil System.

Hydraulic cylinder unit (HCU)

For the HCU, we will concentrate on
two main topics, i.e. the ME control
valves and the exhaust valve actuator
system.

ME control valves
ELVA/ELFI valves
(Curtiss Wright supply)
20 plants with ELVA/ELFI configura-
tion (one control valve for exhaust valve
actuation and another control valve for
fuel injection control) are in service.
For the on/off ELVA valve, a modified
high response valve is undergoing
Fig. 3: 6S40ME-B engine service testing. When this service test-
ing is concluded, the 20 plants will be
updated and service issues with the
ELVA/ELFI configuration will then
be solved.

FIVA Valve (Curtiss Wright version)

The feedback loop of the FIVA valve
position control, Fig. 5, has caused un-
timed injections and untimed exhaust
valve operations owing to different rea-
sons. These reasons are related to the
FIVA valve itself in some cases, and in
other cases to the part of the feedback
loop in the Multi Purpose Controller
(MPC), see Multi Purpose Controller
chapter.

The electronics on the Printed Circuit

Board (PCB) in the Curtiss Wright FIVA
valve showed, in the original version,
Fig. 4: LNG carrier with 2 x 6S70ME-C engines thermal instability causing untimed ac-

4
 pilot slide should be moved downward,
FIVA Valve
thereby increasing pressure on the
CCU top of the main slide and moving the
main slide downward. This will result in
exhaust valve opening or stop of fuel
MPC 4-20 mA LVDT
Electronic
injection.
Requested position
(table) synchronised ÷ When the pilot slide is moved upward
to crankshaft position + PD ±9V Current ±9A 6
(Fig. 8, right hand side), pressure on
control amplifier
the top of the main slide is decreasing
Slide pos. [mm]
5 and the main slide is moved upward
+6,3

0
enabling closing of the exhaust valve or
-7 fuel injection. If a drilling chip is stuck
Exhaust valve opening Injection AC.
in between the pilot and the main slide
when the exhaust valve is closing, there
is a risk of fuel injection just after clos-
ing of the exhaust valve. This will create
Fig. 5: FIVA valve position control an excessive pressure build-up in the
combustion chamber and there will be
a risk of cylinder cover lifting. This was
tions of the valve. The reason was an Fig. 8 gives an explanation of what hap- the reason for the two cylinder cover
analogue voltage regulator generating pens if a loose drilling chip is stuck be- lifts on the 6S70ME-C engines at test-
an excessive amount of heat raising tween the pilot slide and the main slide. bed in 2007.
by itself the temperature by 35ºC on
the PCB. In some cases, this caused Fig. 8 (left hand side) shows the valve in FIVA Valve (MAN B&W version)
a temperature shutdown of the LVDT balance. This means that the constant During 2007, the first vessels with MAN
converter in the feedback loop, result- pressure on the bottom of the main B&W FIVA valves controlling ME engines
ing in the above-described unstable slide is balanced by a pressure creating went into service.
function of the FIVA valve. The solution a similar force in downward direction,
was to exchange the analogue voltage thus keeping the slide in neutral (“zero”) The MAN B&W FIVA valve can be seen
regulator with a switch mode regulator, position. in Fig. 9. It consists of a valve main
Fig. 6. Hereby, the temperature of the body on which the Parker pilot valve
PCB was lowered by approx. 35ºC. In order to open the exhaust valve or to and the H. F. Jensen feedback sensor
stop fuel injection (Fig. 8, centre), the are mounted.
Furthermore, in order to safeguard
against untimed movement of the FIVA
main slide due to an erroneous feed- New Old
back signal, improved supervision is
introduced by new software, see Multi
Purpose Controller chapter.

During 2007, we experienced a cylinder

cover lift on testbed with 6S70ME-C
engines twice. The reason for these in-
cidents was untimed movement of the
FIVA valve main slide owing to a drilling
chip left inside the main slide during
production, Fig. 7. After discovering this
production mistake, we have, together
with the sub-suppliers, cleaned/re-mac-
hined approx. 500 main slides to avoid Fig. 6: FIVA valve feedback failure: exchange of analogue voltage regulator with switch
loose drilling chips inside the FIVA valves. mode voltage regulator

5
 Loose drilling chip found in valve from unit where cover lift had occurred
Fig. 7: CWAT FIVA valve: chips found in main slide

With respect to the H. F. Jensen feed-

Zero position Increased set point Decreased set point
back sensors, we have experienced
two different problems:

A: Breakage of the rod in the sensor.

The rod in the sensor has been
redesigned, Fig. 11a

B: Broken or loose connection

between the print board and the
external connector (type: Canon),
Valve in balance Exhaust valve open or Exhaust valve closed or Fig. 11b. Redesign of the connec-
stop fuel injection fuel injection tion has solved the problem.

Fig. 8: CWAT FIVA valve: Movement of pilot valve and main slide

For the Parker valve, we have seen a Parker pilot valve

number of units failing because of:

A: Broken bushing for the pilot slide,

Fig. 10a. This item was rectified
during the prototype testing period

B: Earthing failure owing to damage of

a flexible wire strip inside the valve,
Fig. 10b

C: Malfunction owing to failing

operational amplifier, Fig. 10c

Parker valves with certain serial num- HF Jensen

bers have been replaced in service. feedback sensor

Fig. 9: MAN Diesel FIVA valve

6
 Welded disc
Steel/steel welding

μ metal core

Screwed connection
μ metal tube secured by welding
Micro cracks Steel/mu-metal
Steel base
Steel base
Fig. 10a: Broken bushing Steel base

Vibration damping
with O-ring
New version

Old version

New design Old design

Fig. 10b: Damaged wire strip

Fig. 11a: Redesign of rod for sensor in H.F.
Jensen feedback sensor

Fig. 11b: Connector breakdown, H.F.

Jensen feedback sensor

Failing operational amplifier

Fig. 10c: Parker valve

In parallel with solving teething prob-

lems with the Parker pilot valve and the
H. F. Jensen feedback sensor, other
makes of pilot valves and feedback
sensors are being tested.

FIVA Valve (Bosch Rexroth version)

Service tests of the Bosch Rexroth
FIVA valve, Fig. 12, on a 12K98ME have
been concluded successfully. Bosch
Rexroth FIVA valves are now the third
alternative for control valves for the
present ME engine series.

Fig. 12: Bosch Rexroth FIVA

7
Cavitation in the exhaust valve An orifice in the drain line from the FIVA
actuation system valve, Fig. 14, has been introduced to
Cavitation in the exhaust valve actua- reduce the acceleration of the actuator
tion system has been seen in the ex- piston and hereby eliminating cavitation
haust valve top part, Fig. 13a, as well on the actuation side.
as in the exhaust actuator top cover,
Fig. 13b. At the time of writing, we are monitoring
cavitation development after the intro-
Furthermore, damage to the oil inlet duction of the orifice in the FIVA return
non-return valves on the actuator top line. However, in parallel, we are testing
cover indicates excessive pressure further modifications:
fluctuations in the exhaust valve actua-
tion system.

Fig. 13a: Cavitation in exhaust valve top part Fig. 13b: Cavitation in exhaust actuator top cover

Introduction of ring orifice corresponding to a 20 mm orifice in the hydraulic system for exhaust valve on K98ME-C
Fig. 14: Orifice in the drain line from the FIVA valve

8
A: Reduced braking of the exhaust
valve by introduction of an orifice
(small hole) in the damper piston,
Small hole
Fig. 15 in damper
piston to New low pres-
B: Low pressure oil supply in the top of prevent sure supply
the exhaust valve, Fig. 15. It is con- cavitation through non-
return valve
sidered to move one of the low
pressure supplies at the actuator,
Fig. 16, to the top of the exhaust valve

Whether (A) or (B) or both (A) and (B)

will be needed in order to completely
eliminate cavitation damage will be de-
cided in the coming months.

Fig. 15: Exhaust actuation system, scheduled test rig tests

Fig. 16: Non-return valves

9
Multi purpose controller (MPC)
In 2007, we experienced one severe case
of cylinder cover lift on a 6S60ME-C
engine in service. After investigations
into the parts involved on the cylinder
unit in question, it was concluded that
the reason was an error in the feedback
loop for the FIVA control, Fig. 5. However,
in this case the error was in the MPC
part of the feedback loop. A loose/bro-
ken connection, Fig. 17, in the feedback
circuit of the MPC was found during the
investigations.

Countermeasures in this relation were

divided into three parts.

Firstly, a circular letter warning against

a specific alarm sequence was issued in
order to exchange MPC’s with similar
potential defects. This circular letter
was sent to all operators of ME engines.
Fig. 17: Broken/loose connection in the MPC
Secondly, it was concluded from tests
that the reason why the error in the
feedback loop of the MPC caused an
untimed injection was that the feedback
signal froze on a low value. When the
closed loop control tried to rectify posi-
tion of the main slide in the FIVA valve, 18,66 18,25 Old feedback
it moved the main slide towards fuel signal
injection and continued to do so until
injection (untimed) was accomplished. 5,75 5,33
New feedback
This was done because of the frozen
signal
low value feedback signal.

Based on this knowledge, it was decided

to invert the feedback signal, Fig. 18.
By doing this, a frozen low value feed-
back signal will result in FIVA main slide
movement towards untimed opening of
the exhaust valve. This is considered to
be “failing into safe mode”.

Thirdly, in order to safeguard further Fig. 18: FIVA valve flow area diagram
against similar incidents, a new soft-
ware version with closer supervision of
the feedback signal [1] and additional All ME engines in service have been or
supervision of the fuel plunger move- will be updated with the above counter-
ment [2] has been introduced, Fig. 19. measures.
Control processes including the super-
visions [1] and [2] are seen in Fig. 20.

10
 New software FIVA Valve

CCU

2
MPC 4-20 mA LVDT
1 Electronic
Requested
position (table) ÷ 6
synchronised to + PD ± 9 V Current ±9A
crankshaft control amplifier
position
Slide pos .[mm] 5
+6,3

-7

Exhaust valve opening Injection

C.A.

Fig. 19: FIVA valve position control

Fig. 20: Crankshaft related control processes

11
Hydraulic power supply (HPS) The fault has been
In 2006, we experienced a break- identified to the design of
down of the bearings in an HPS on a the first S70ME-C plants.
12K98ME engine. A design review was
initiated and modifications were imple- All plants with this design
mented, both for new engines and for will be modified with
engines in service. After this incident double amount of screws
we have not seen further incidents re- in pos. 7 a high friction
lating to the bearing/bushing design on disc between pos. 2 and 3
the ME engines. and bigger dimension on
pos. 6
For a series of the first S70ME-C plants
the assembly of the chain wheel and
the gear wheel to the common shaft
has shown to be under-dimensioned,
Fig. 21. Upgrade of the bolt connec-
tions has been introduced to the con-
cerned vessels in service.
Damaged screws
Servo oil system
For the present ME engines two Fig. 21: S70ME-C assembly of shaft for hydraulic pump drive
alternatives are available as servo oil
system:

A: A standard system where engine

system oil is processed through a
6-10 μm full-flow fine filter and then
led into the hydraulic pumps of the
HPS.

B: An optional system where a separate

servo oil system with a separate
tank system is used, Fig. 22. In this
case, the oil is being cleaned by a
cleaning unit (filter or purifier) mount-
ed on the separate oil tank.

During 2007, we experienced one case

of severe contamination of the servo oil
system owing to breakdown of the ship
side transfer pumps. This happened on
an installation equipped with a separate
servo oil system (B).
Fig. 22: Separate hydraulic oil system
Apparently, the pumps of the screw
type produced the contaminating prod- system, Fig. 24. The important change add a ‘‘water-in-the-oil-monitor’’, con-
ucts rather quickly and therefore, a lot is that a 6 μm full-flow fine filter has nect the oil temperature measurement
of debris, Fig. 23, ended up inside the been introduced, also on the separate to the alarm system and install a metal
ME control valves. servo oil system. detector just before the hydraulic power
supply.
After this incident, we have revised our For engines in service with a separate
specification for the separate servo oil servo oil system, we recommend to

12
 Conclusion
During 2007, we solved and concluded
several service-related issues for the
ME/ME-C engine series. This work will
naturally continue in 2008 and the main
focus is still to make the updates in
service without disturbing the operation
of the vessels. This has been possible
in most cases in 2007 and we are con-
fident that this will also be the case in
the future.

Fig. 23: Breakdown of transfer pumps

Fig. 24: Separate hydraulic oil system, updated version

13