EE Twelve cylinder diesels and the EAR 90 class

[John Ashworth]

Posted on the East_African_Steam Yahoo group

The English Electric (EE) 12-cylinder models present an interesting
case study, in particular because from 1959, there was a sequence of
Australian models as well as those from the UK parent company. The
respective development pathways, whilst not completely independent,
were somewhat different.

The "twelve" was actually the last of the original EE RK/V engine
variants to be built for a production order. The 4-, 6-, 8- and 16-
cylinder models had all been produced in Mark I form for a variety of
locomotive and railcar applications. There were no railway
applications for the 12SVT Mark I engine, but it was the first Mark
II model to be series-built, in 1953. It was also the first to be
built in charge air-cooled form, as the 12CSVT, in 1960. At the tail
end of EE locomotive production in Australia were built a small
number powered by the 12CSVT Mk III engine. However Australian
production had used a modified version of the Mark II engine since
1964.

EE generally designed locomotives to meet individual railway
requirements, using standard components. Nevertheless, established
designs were often adapted for other customers, and this is apparent
in the 12-cylinder sequence, which accordingly could be described as
being a series of quasi-standard locomotives. A greater degree of
standardization is apparent in EE's sequence of 6-cylinder models,
though.

The list of EE 12-cylinder models, in chronological order of first
appearance, is:

1. Queensland Railways (QR) 1200 class, 10 built.
2. New Zealand Railways (NZR) Df class, 10 built
3. Malayan Railways (KTM) 20 class, 26 built
4. QR 1250 class, the first EE Australia variant, 17 built
5. Sudan Railways 1000 class, 65 built
6. East African Railways 90 (later 87) class, 44 built
7. British Railways (BR) 37 class, 309 built
8. Western Australian Government Railways (WAGR) C class, 3 built
9. Rhodesian Railways DE3 class, 16 built
10. QR 1270 class, 30 built
11. WAGR K class, also Goldsworthy Mining A class, 17 built
12. QR 1300 class, 45 built
13. WAGR R and RA classes, 18 built
14. AIS D34 class, 1 built
15. Ghana Railways & Harbours 1851 class, 16 built
16. Tasmanian Government Railways (TGR) Z class, 4 built
17. QE 2350 class, 16 built
18. TGR ZA class, 6 built

By my calculations, the total number of EE 12-cylinder locomotives
built was thus 653, of which 496 were of UK origin, and 157 came from
Australia.

This total comfortably exceeds the production of locomotives powered
by other RK/V engine variants, again by my calculations as follows:

4SRKT 109 (locomotives only; railcars excluded)
6SRKT 249 (excludes the 6KT engine)
8SRKT & 8SVT 408 (59 8SRKT & 349 8SVT)
16SVT 277

Additional statistics about the EE 12-cylinder locomotives are that
309 were for use in the UK, EE's home country, and 344 for "export"
markets (relative to the UK.) 327 were standard gauge, 256 were Cape
gauge and 70 were metre gauge; there being no broad gauge examples.
As built, 190 were equipped for hauling air-braked trains and 445 for
hauling vacuum-braked trains, whilst 18 were dual-braked. Many
received subsequent braking system modifications, though. The high
number originally equipped to haul vacuum-braked trains is, perhaps,
a reflection of the British-influenced markets in which EE traded.
All had six traction motors, and the majority, 583, were of the Co-Co
wheel arrangement. 60 were of the 1-Co-Co-1 type, and 10 were 2-Co-
Co-2. 96 had 12SVT Mk II engines, 535 had 12CSVT Mk II engines and
22 had the 12CSVT Mk III type. As far as I know, 115 of the 12-
cylinder Mk II engines were of the Australian special variant, 30
12SVT and 85 12CSVT. 20 locomotives had 10-notch electromagnetic
control systems, the balance (this needs to be confirmed) had
electropneumatic control systems with continuously variable pneumatic
throttles, which was the EE standard fitment from the mid-1950s, and
which formed the basis for the BR `blue star' MU coupling class,
although the latter also made provision for (optional) electrically
controlled starting notches, not used by EE. 202 of the locomotives
were fitted with dynamic braking equipment. All but 10 were fitted
for multiple unit operation as built.

BR, with 309 of its 37 class, was the biggest user of EE 12-cylinder
models. Next came QR, with a total of 118 spread over 5 basic
models, although there were subvariants. QR was also the first
operator to buy 12-cylinder EE locomotives, so are more detailed
study of the group logically starts with the QR 1200 class, which
then conveniently links to both the later UK and the Australian
models. At least basic information on most classes is reasonably
available, and of course the BR 37 has been the subject of many
treatments in the literature.

The EARH 90 class was the second English Electric 12CSVT-engined
model to be delivered, but the first to be ordered. The initial
order, for 8 units was announced in October 1958, and an increase to
10 units was announced in March 1959. This was EARH's first order
for line-service diesel locomotives. A 13.5 ton maximum axle loading
was imposed, to enable the locomotives to work northwest of Nairobi
to Nakuru and Kampala, as well as between Mombasa and Nairobi, which
section alone would have allowed a higher axle loading. This axle
loading constraint required a multiaxle design, as it is unlikely
that EE could have built a compliant 12-cylinder Co-Co model.
Unsurprisingly, EE used a 1-Co-Co-1 wheel arrangement. The resulting
locomotive was largely a new design, although it included features
drawn from the QR 1250 class (body style and general layout) and the
Rhodesian Railways (RR) 16-cylinder DE2 class (running gear and in-
frame fuel tank). What it was not, though, was simply a 1-Co-Co-1
variant of the QR 1250 with 12CSVT in place of 12SVT engine.

Nevertheless, the QR 1250 makes a useful yardstick for comparison
purposes. The EARH 90, at 51'0" over headstocks, was a little longer
than the QR 1250, at 49'6". This extra length was most likely
required to accommodate the more complex running gear, although it
probably also gave a bit more space to accommodate the dynamic
braking unit and a higher capacity cooling group. The total
wheelbase was 41'6", as compared with 40'0" for the QR 1250. The
equipment layout for the most part followed established EE practice.
The nose compartment housed the leading bogie traction motor blower,
which was motor-driven. The T-shaped main equipment cubicle was
immediately behind the cab; then came the dynamic braking unit, which
was mounted high – just below the cantrail - with a crosswise
orientation, fanshaft horizontal. Then came the generators, the
engine, followed by the radiator compartment with mechanically-driven
vertical-shaft fan, and finally the rear-compartment, housing the air
compressor, mechanically-driven from the radiator fan gearbox, and
the trailing bogie traction motor blower.

The running gear was based upon that of the RR DE2, which had proved
successful in service. Thus, the bogie frames were one-piece
castings by Henricot. Because the EARH 90 was shorter, the axle
spacings were all reduced by 6 inches. Each bogie had an overall
wheelbase of 17'6", with a rigid wheelbase of 12'0" equally
distributed, and the two inner bogie axles were separated by 6'6".
The pivot centres, placed between the pilot and outer driving axles,
were 37'10" apart. Consistent with EE's thinking about maximizing
the advantages obtainable from the 1-Co-Co-1 wheel arrangement where
it was necessary to use same, the axle spacings were chosen to obtain
maximum bending moment relief, so reducing vertical railhead forces.
As this required relatively close coupling of the bogies, a
conventional suspended fuel tank was precluded, hence the use of an
in-frame fuel tank, an EE feature that went back at least as far as
the New Zealand Railways De class. The main bogies were
interconnected by a lateral spring control mechanism that helped
ensure optimum wheel flange angles in curves, so reducing lateral
railhead forces. One way of looking at this is that the coupling
allowed the leading bogie to pilot the trailing bogie into curves,
the leading bogie itself being guided by its own pilot truck. Wheel
diameters were the same as on the DE2, namely 28½" pilot and 37½"
driving. The main bogies had three-point load transfer from the
mainframe, with a resiliently mounted pivot between the pilot axle
and the outer driving axle, and a pair of coil spring bearers between
the centre and inner driving axles. Equalization for each bogie was
continuous from pilot truck axle to inner driving axle. This was a
change from the DE2 bogie, which was equalized in two groups,
although the casting did make provision for full equalization should
it have been required. As with the DE2, the driving axle springs
were of the leaf type interconnected by equalizing bars, but at the
fixed attachment points, rubber bushes were used in place of the
auxiliary coil springs used on the DE2. All three traction motors on
each bogie faced outwards, consistent with high-adhesion bogie
practice. EE claimed that this bogie design virtually eliminated
intrabogie weight transfer, whilst the wide pivot spacing minimized
interbogie weight transfer. I have never seen the benefit quantified
in the same way that EE Australia did for its high-adhesion Co bogie
first used on the Western Australia Government Railways R class, but
taking the latter as indicative, EE's view could well have been that
a 1-Co-Co-1 locomotive with 81 tons adhesive weight would for
practical purposes match a similarly-powered 90 ton Co-Co with more-
or-less conventional bogies, whilst offering lower dynamic railhead
forces as well as (fairly obviously) lower static railhead forces.
EE certainly made much of the capabilities of it own-design 1-Co-Co-1
running gear (here one needs to be careful to exclude the non-EE
design bogie that it was forced to use, against its better judgement,
for the British Rail 40 class) and noted that its good performance
had been verified by the railway administrations using it.

The 12CSVT Mk II engine had three manually adjustable governor power
settings that allowed optimization for altitude, bearing in mind that
the route embraced the range from sea level to 9136 ft elevation.
1840 hp (gross) was available up to 5500 ft, 1800 hp up to 7800 ft,
and 1775 hp up to 9136 ft. One assumes that the settings were chosen
according to which part of the EARH system the locomotives were
assigned. The main generator was the EE822 model, and the six
traction motors were the new EE537 4-pole model, connected in
permanent series-parallel (2S3P) with two stages of field weakening.
I have not been able to verify the gear ratio, but most probable was
72:15, fairly standard for the EE537 motor. The 45 mile/h maximum
service speed would not have required faster gearing. The overhung
auxiliary generator, model number unknown, was of 48 kW capacity.

The EARH 90 was fitted with the by-now standard EE air-throttle
control system, with the EE governor and the new hydraulically
operated load regulator. The two driving stations were fitted with
EE's then-standard two-lever control stands. The throttle lever also
operated the dynamic brake according to the standard EE 3-notch
protocol. The main driving station was on the right hand side.
There was a second driving station on the left-hand side, but this
was not diagonally opposite and reversed as might have been
expected. Rather it seems to have been arranged to allow
bidirectional operation during shunting operations.

The braking system was air for the locomotive and train, EARH being
an air-braked road. However, the design made provision for the
retrofitting of vacuum train brake equipment if required. At the
time, it was evidently still thought possible that the EARH system
would be converted from metre to Cape gauge to align with the rest of
Southern Africa. The same conversion would also have required a
(retrograde) change from air to vacuum brakes, the latter being the
Southern African standard, with at least a period of dual-braking
capability being required. One can wonder how vacuum brakes would
have performed at 9000 ft altitude. Also, it is not immediately
apparent as to where the vacuum exhausters would have been
accommodated on the 90 class, bearing in mind that both the large
compressor and the dynamic braking equipment would have been
retained. Perhaps EE was thinking in terms of using a combined
exhauster-compressor unit in place of the air compressor.

I have not been able to find definitive information about the type of
air braking system fitted to the EARH 90 class, other than that the
initial batch had Westinghouse UK equipment. Had an American-
type "schedule" system been fitted, most likely it would have been
noted in the trade press descriptions. So more likely is that the
braking system followed British precepts, with physically separate
driver's valves for independent and train brake control, and an
electrically operated independent-release-after-automatic-application
function. Certainly EARH did not have a history of using schedule
systems on its steam locomotives, late examples of which were fitted
with Westinghouse No. 4 automatic brake valves and Gresham & Craven
Mk IV locomotive steam brake valves. The layout diagrams show that
the two driver's brake valves are somewhat separated, with that for
the automatic brake being to the driver's right, and that for the
independent brake a little to the left, ahead of the control stand.
Clasp brakes were fitted to the driving wheels, withy one brake
cylinder per wheel. The pony truck wheels were unbraked.

Another unknown is the electrical capacity of the dynamic brake
unit. For the second series, the peak braking effort is shown as 30
000 lbf at approximately 21 mile/h, which suggests around 1200 kW.
The second series is said to have had a greater dynamic braking range
than the first series, so it is possible the latter had a smaller
capacity unit. As described, it is stated that the locomotive brake
is interlocked with the dynamic brake so that both cannot be applied
simultaneously. A literal interpretation suggests that the 90 was
fitted with a conventional lockout system. But if so, it was a
departure from established EE practice. In its previous diesel-
electric locomotive dynamic braking installations, EE had used an
anti-compounding system, in which the dynamic brake was released if
locomotive brake cylinder pressure reached a predetermined level,
typically 23 lbf/in².

The EAR 90 was equipped for multiple unit operation. There was a
central EE elbow-style jumper socket at each end on the front sheet,
along with three "plug-in" hose connections arranged in a triangle.
Most probably these were for respectively main reservoir, engine
speed control and independent brake. The 90 was MU compatible with
the later and smaller EE-built 91 (71) and 72 classes, but beyond
that EAR did not seem to be concerned to establish a single common MU
standard. Later diesel locomotives from other builders were equipped
with control and MU systems that were more-or-less their respective
builder's standards.

Notwithstanding the 13.5 tons axle loading specification, the first
series were built to a slightly lower 12.8 tons number, giving an
adhesive weight of 76.8 tons. The total weight was 97.5 tons. The
continuous tractive effort is consistently quoted as 44 500 lbf,
although there is some variety in the corresponding minimum
continuous speed, which is variously reported as 11.5, 11.7 and 12¼
mile/h. The top speed is usually reported as 45 mile/h, but this
would have been a track limited speed, as the expected 72:15 gearing
would have allowed 60 mile/h, and there is no reason why the running
gear would not have accommodated this on suitable track.

Written by Steve Palmano

[John Ashworth]

I'd be very interested if anyone has any details and/or photos of the Sudan Railways 1000 class, as mentioned in #5 above.

[Ian Roberts]

I did see and photograph one of these locos in Port Sudan in 1960, sadly there was extreme camera shake and the slide was discarded.
The loco was in a blue and yellow livery and appeared to be in ex works condition.

[John Ashworth]

2nd instalment posted on the East_African_Steam Yahoo group

Continuing the English Electric Twelve Cylinder model series, this is the second
of two parts on the East African Railways 90 (later 87) Class.

EAR placed three repeat orders for the 90 class. A second batch of 14 was
ordered in April, 1963 for delivery in the first part of 1964. The order for a
third batch of 12 was reported in Railway Gazette for October 01, 1965, and was
reported as being in process of fulfilment in English Electric Journal for
January-February 1967. The order for the fourth and final batch of 8 was
reported in EE Journal for June-July 1967, and is understood to have been
delivered during 1968. The total built was 44.

Some changes were made with the second batch. The driving axle load was
increased from 12.8 to 13.5 tons, this being the result of the confirmed good
riding properties and low track stresses of the first batch. Total weight was
increased to 101 tons, of which 81 tons was adhesive.

Full power, 1840 hp site gross, was available from sea level up to maximum
altitude, 9136 ft, so that no governor adjustment was required with change of
altitude. This was attributed to the use of an optimized and matched
turbocharger design. Possibly there was a change of governor, from the EE type
to the Woodward PG, as well, but if so this is not recorded. By then the PG was
available with inbuilt altitude compensation, which might have been desirable
under the circumstances. The circumstantial evidence is fairly strong, as there
is the precedent of the Rhodesian Railways DE3 class and the later example of
EAR's own 91 class, both being fitted with the Woodward PG.

Main generator and traction motors were the same as for the first batch. As
mentioned in the first part, the second batch had a dynamic braking system with
a greater range; it has also been described as having more capacity. One would
need to see the curves to know one way or the other.

Metcalfe-Oerlikon driver's brake valves and distributor were fitted in place of
the Westinghouse valves of the first batch. Davies & Metcalfe apparently did
quite a bit of business with EAR in the 1960s.

Retained across all four batches were the Stones headlamps. The same had been
the case with the Sudan Railways 1000 class. One may surmise that these may
have been preferred because of lower replacements costs.

Circa 1970, EAR implemented a new locomotive classification system, still using
two digits, but with the first digit indicating size/power output. Thus the 90
class became the 87 class.

As far as I know, Kenya Railways (KR) received most of the 87 class following
the breakup of EAR in 1978. One source indicates that it had 35. Whether some
or all of the others were eventually transferred to KR I do not know.

One may reasonably deduce that EAR was happy with these locomotives in general,
and in particular with the tracking, riding and adhesion properties of the
specific 1-Co-Co-1 wheel arrangement. The latter is evident from the fact that
it chose the 1-Bo-Bo-1 wheel arrangement for its 91 class light diesel-electric
locomotives, also built by EE. Bidders for this business were asked to offer
any/all of the wheel arrangements Co-Co, 1Bo-Bo1, 1-Bo-Bo-1 and 1-Co-Co-1.
1-Bo-Bo-1 was evidently chosen because it offered the benefits of the pilot
truck arrangement without undue complication, four motors being considered
adequate for the duties envisaged.

For its next generation of medium and heavy locomotives, EAR turned to MLW.
This may have been for financial reasons as much or even more than for technical
reasons, as the purchase was backed by a Canadian Government loan, interest-free
over 40 years, with a 10-year moratorium. That is perhaps why the EAR MLWs were
fitted with Canadian GE electrical equipment, whereas one might have expected
AEI equipment in an Alco or MLW destined for a Commonwealth country. MLW was
apparently anxious to break into the African market, and to design for it
accordingly.

Thus there was EE influence in the form of the 1-Co-Co-1 wheel arrangement,
which was used for the EAR MLW-built 88 and 92 classes ordered in 1970. The 88
was more-or-less a direct successor to the 90 (by then renamed as the 87 class),
and the 13 ton axle loading specified for it meant that more than 6 axles were
required, so that 1-Co-Co-1 was a logical choice anyway. The more powerful 92
class, with a 16¼ ton axle loading, could have been built as a Co-Co, but EAR
chose to have it with the 1-Co-Co-1 wheel arrangement. Maximum commonality with
the 88 was no doubt a factor here, but so would have been EAR's preference for
the pilot truck arrangement. MLW used a quite different bogie design, and did
not use an interbogie coupling. Nor were all of the longitudinal wheel spacings
arranged to maximize vertical railhead stress relief. So in the rail stress
department, the MLWs were probably not quite as good as the EEs. Possibly the
13 ton axle loading limit for the 88, as compared with 13.5 tons for the later
batches of the 90 (87), was set in deference to that expectation.

At around the same time Nigerian Railways became interested in the 1-Co-Co-1
wheel arrangement, and it seems very likely that this stemmed from its awareness
of the EAR experience with its EE fleet. It acquired both Hitachi and MLW
locomotives with this wheel arrangement. The Hitachi model was part of a family
that included the Sudan Railways 1400 class, which as previously recorded, was
influenced by EE practice in other directions.

Had EE been able to maintain a supplier position with EAR, then one may suppose
that it would have supplied more of the 87 class, and perhaps a larger version
of 2500 hp or so. That would have been possible using the 12CSVT Mk III engine,
possibly with an alternator rather than a generator. Such a locomotive might
have been longer, for example to accommodate a larger cooling group, and perhaps
it would have adopted the RR DE2 dimensions. Probably it would have had a
rearranged internal layout with mechanical auxiliary drives, as discussed in
connection with the QR 1270 class. EAR had already experienced that change with
its 90 class, which had a "mechanical" layout.

Looking back to the late 1950s when the 90 class was ordered, this was at about
the same time that South African Railways ordered its GE U18C1 1-Co-Co-1 fleet
for use in South West Africa. One may wonder whether EE bid for the South
African business, and if so was it with a locomotive very much like the EAR 90,
and did GE bid for the EAR business. EE had supplied DC electric locomotives to
SAR in the 1950s, and this, coupled with SAR's observations of the RR DE2 fleet
would have made it a credible supplier, but SAR seemed to be predisposed towards
American diesel locomotives, and already had a GE fleet in service, so GE
probably had the inside running there. Also, it would have been better placed
to offer relatively quick delivery of the large (115) number required. On the
other hand, supply from the UK was probably preferred for EAR. However, I do
know that Cockerill of Belgium bid 1-Co-Co-1 designs against both the EAR and
SAR requirements.

Some trade publication references for the EAR 90 are:

The Railway Gazette, September 02, 1960, "Diesel Locomotives for East African
Railways", p.280ff.

The Railway Gazette, May 15, 1964, "Large Diesels for East Africa", p.389ff.
This was reprinted as an EE brochure, then reissued by GEC.

In closing, I'll nominate the EAR 80 (87) class as the archetypal EE
twelve-cylinder model. My reasons are several, as follows:

1. It was a multiaxle design using the 1-Co-Co-1 wheel arrangement, which
although not unique to EE, was an EE specialty.
2. It had the cab/hood body style, again not unique to EE, but one that EE made
its own.
3. It had the classic EE equipment layout, including motor-driven traction motor
blowers, overhung auxiliary generator, mechanically driven fan and compressor.
4. It was designed for use in difficult conditions, sea level to over 9000 ft
altitude, tropical to temperate ambient, limited axle loading.
5. It had the EE standard air throttle control system. It also had dynamic
brakes, which whilst not a determining factor, was nevertheless a significant EE
feature in the export realm.
6. A reasonable number were built over an extended period, and there was a
derivative model, namely the RR DE3.
7. It was by all accounts a successful design with a long service life. (Some
might still be operated by KR.)

As can be seen, most of the determining parameters are a measure of "EE-ness",
as it were, and not elements of some kind of overall merit ranking, which is not
the intent here.

Steve Palmano

[John Ashworth]

Some might still be operated by KR

I can confirm that some still are being operated by KR.