St Brides Pump House History and Restoration

For many years, the pump engine was maintained by the late Lionel White (pictured below), who was made an Honorary Member of the Friends in appreciation of his outstanding contribution. Lionel also ran his own website which documented the renovation work done by the Friends and National Park in 2000-2001, and the valiant efforts to get the pump engine running again. This information is now available here, with thanks to Lionel’s family.

Back to main page


History of the pump house
Restoration of the pump house 2000-2001
Restoration of the pump 2001-2002
Facts and figures

History of the pump house

During the years 1905–1913, large extensions were added to St Brides Castle for the Sixth Baron, Lord Kensington, the work being undertaken by James Barbour and Bowie of Dumfries. ‘Building News’ 1913 reports that ‘a new water supply, drainage and electric light installation forms part of the scheme thus completed’. The ‘Pumping Power House’ was fitted with Tangye engine and pumps and a pipeline brought water from a spring in the east under the Dale road to an underground reservoir on the north side of the Pump-house. The reservoir was a brick and concrete construction, about 16 x 20 feet, 8-9 feet deep containing about 16,000 gallons of water. The pipeline then followed the churchyard wall on the seaward side, crossing the grounds of the Castle and the Drive, to the road leading to Home Farm. From the corner of the Castle it rose to the reservoir on the hill to the south-west of the Castle, the water having been pumped from the pump house, near sealevel, to a height of 185 feet (57m). Through a parallel water-main the water for St Brides Castle came ‘by gravitation to a cistern on the roof. There was a Filter House for filtering the water to the Castle, laundry, stable-yard and fire-plugs surrounding the building’.

Other branches of the pipeline supplied the garden and Home Farm. Before the sale of the property in 1922, St Brides Rectory was connected to the main. This was by an agreement between the Sixth Baron and the Representative Body of the Church in Wales, signed in January 1921. When the Castle was officially opened as a Hospital on June 12th 1923 it too was to benefit from a good supply of fresh water piped from the Pump-house.

The Engine and Pump remained in service until the late 1960’s.

(Selected extracts from the booklet‚ ‘St Brides, Pembrokeshire’ by Dorothy Willcock)

Back to top

Restoration of the pump house 2000-2001

The Pump-house is situated within a small garden at St Brides Haven, close to St Brides Church and the former Rectory, and was built to provide St Brides Castle with fresh water. It is now owned by Pembrokeshire Coast National Park Authority, but the Friends of Pembrokeshire Coast National Park were interested in restoring the building and bringing the machinery into full working order. Bob Powell (founder member of the Friends with the late Claude Parry} had a vision of the Pump-house restored to its former glory, with the oil engine and water pump working on demonstration days. The larger room, which used to house the main pump, would have a sequence of illustrated display panels with the history of the Kensington Estate and general information on the National Park. There would be fixed seating to provide a wet-weather refuge for rest and refreshment. His enthusiasm for the Project was so infectious that he quickly gathered around him a team of Friends who were inspired by his views. Early in the year 2000 Bob announced that the Project had been agreed with the National Park Officer, and the required approvals and finance had been received via the Landfill Tax Credit Scheme, thanks to the Texaco Oil Refining Company and the Government’s Regulator (ENTRUST).

Thus it was that the first work-party of Friends and Staff of the National Park assembled on March 7th 2000 and enjoyed some hours of very hard work, cutting back foliage and overhanging branches before clearing the interior of the Pumphouse. It was then that the reservoir of clear water was revealed through a small door at floor level, to the amazement of the workers on that first day at St Brides, and was photographed in detail by Dr Phil Robinson. After this the work-party met every other Tuesday, then every Tuesday. for the next year and this involved many skills – carpentry, decorating, engineering, electrical engineering, gardening, historical research, artwork, labouring and tea-making. The workers were Bob Powell, Brian Meopham, Doug Voss, Arthur Bebbington, Graham Morgan, Roy Rothwell, Derek Rowland, John Ratcliffe, Myles Huthwaite, Colin Thomas and Lionel White. In October 2000 Bob became seriously ill and Brian Meopham took over as ‘Clerk of Works’.

It was hoped that the Pump-house Project would be completed for Easter 2001 in time to celebrate the tenth anniversary of the formation of the Friends of Pembrokeshire National Park. The AGM was planned for Saturday April 28th 2001 at St Brides Castle and after much hard work everything was ready for the opening of the Pump-house that afternoon. It was just as this ceremony was starting that the news was brought of Bob Powell’s death at Withybush Hospital, Haverfordwest. All present stood in silence as a mark of respect to a man who had made such an enormous contribution to the Friends and to the National Park as a whole for many years.

Chairman John Ratcliffe welcomed members and guests who included Councillor Cawood, Nic Wheeler, National Park Chief Executive, and David Harries from Texaco’s Pembroke Plant, without whose support the Project would not have been possible.

The Pump-house was then officially opened by Gordon Cawood, Chairman of the Pembrokeshire Coast National Park Authority, who praised the restoration work on the Pump-house and the transformation of a derelict property to what is virtually a new Information Centre for the St Brides area. Thanks were also due to Roger Hiscock and Brian Southern of the National Park’s Graphic Services Department who coordinated the preparation of the display panels illustrated by Jean Pugh and whose advice and professionalism were vital to the Project.

Since the opening Lionel White, supported by Arthur Bebbington, has continued his work on the engine and pump, acquiring and fitting several replacement parts. In October 2003 Lionel reported that the engine had been run on twelve separate occasions during the year with a total running time in excess of eight hours but that work was still required on the pump.

[Professor John Ratcliffe (former chair of The Friends of Pembrokeshire National Park) also wrote an article about the restoration which was published in the December 2002 edition of Pembrokeshire Life ISSN 1354-9332.]

Back to top

Restoration of the pump engine 2001-2002

Following on from the work to restore the fabric of the St. Bride’s pump house, attention focused on the engine and pump. After a general clean up the engine had been persuaded to start and provide background sound for the formal re-opening in April 2001, but it became more and more difficult to start, and the running was very rough. By the end of July, it just would not perform.

The initial fault was a complete lack of a spark. Tests on the magneto suggested an internal short circuit. An internet search identified a firm in Bideford that specialises in the servicing and repair of such items and our magneto was dispatched for evaluation and a formal quote for a complete overhaul. The quote seemed reasonable and the work was authorised with a completion date of 23 November.

It was therefore mid-December before we got round to re-installing the magneto on the engine. The first thing was to check if we now had a spark. The engine was turned over but no spark. Not sure why, but the engine was then rotated in the opposite direction to that in which we had previously run it and produced a good strong spark. A lengthy discussion then ensued as to exactly which way the engine should be run. The two members who had had experiences of the engine during its working life could not agree and we could not get it to fire up in the clockwise direction (previous running being in the anticlockwise direction).

From various sources we had established that the magneto had been made by Hills Brothers of Bristol and was a model HBR1. An exchange of correspondence with the magneto expert resulted in the supply of a reproduction leaflet relating to a Bosch Type “22”, of which the HBR1 is a direct copy. Reading through the booklet it was established that the magneto could be configured to give a spark for either clockwise or anticlockwise rotating engines.

Our magneto was in fact configured for use in the anticlockwise direction, and examination of before and after photographs confirmed that this was the same as before the overhaul. Our diagnosis was that lack of a spark in the required direction was because the points had not been correctly adjusted. The spark in the reverse direction was spurious. The magneto was returned to the expert for re-adjustment at no expense to ourselves.

This still did not resolve the problem and we found ourselves in the situation where the expert was claiming that “the magneto is in perfect working order” and we were saying “It doesn’t work”. The only way that this situation was resolved was by a visit to Bideford (a round trip of 440 miles) to witness testing of the magneto on the in-house test rig. The day before the visit a phone call was received to say that their normal test rig had failed so they could only do a qualitative test rather confirm proper operation but this was enough to establish that we needed to look further at our engine

Picture 1 shows the magneto operating mechanism. The vertical arm is either pulled to the right or pushed to the left for clockwise or anticlockwise engine operation. The picture shows the lever being pushed, further movement to the left causes the lever to “trip” and return to the centre (rest) position by the action of the spring. The spring allows the lever to overshoot the vertical rest position and it is during this overshoot that the points are operated. A wire coil is linked to the lever and moves in a magnetic field. This is what generates the electricity to produce the spark. The critical factors are that the coil must move fast enough to generate the required current and also to cause sufficient overshoot for the points to operate.

The paper scale was a temporary measure and was used to determine what deflections were needed to generate sparks and what deflections were produced by the engine before the mechanism tripped.

The Bosch booklet implied that a spark should be produced with deflections of 30 degrees. In practice a spark was produced with a deflection of 22 degrees. Unfortunately the engine was only producing a deflection of 12 degrees for anticlockwise rotation. It was noted however that if the engine was rotated in the clockwise direction then a deflection of 25 degrees was produced.

With the help of a few measurements and a bit of geometry, figure 1 was produced to give a better understanding of the operation of the magneto drive mechanism.

Careful examination (not possible in the scale of this reproduction) seemed to indicate that the mechanism was biased to give a greater deflection when pulling the lever to the right (clockwise engine rotation) than when pushing to the left (anticlockwise rotation), so we were back once again questioning exactly which way the engine should be run.

Although we had the original cranking handle, this was no help as the head is reversible, reputedly so that it can be used both for the engine and the pump. The pump itself, being a triple ram, should work in either direction. It is just that the gearing between the two causes opposite rotations at the cranking points. The only real clues are in the exact timing of the exhaust valve and of the ignition. The inlet valve is controlled by the pressure/vacuum in the cylinder. The first attempt to establish timing diagrams was inconclusive as we could not establish TDC “Top Dead Centre – the point where the piston is at its highest on the firing stroke” which is the standard reference point for engine timing.

We needed a slice of good fortune and it came on April 21st in the guise of Marc, Nick and Sally, members of the Wychavon Oil Engine and Preservation Society from Evesham Worcestershire. They were aware of our Pump House project and could not wait to see the engine close-up, especially young Marc. After a lengthy discussion about the problems that we had been having Marc got to work. He determined the TDC (apparently if there is no specific marking then there is normally a mechanical feature such as the crankshaft / flywheel key-way which in our case is at its lowest point at TDC.) It was then a relatively simple process to determine that the engine was set-up to run anticlockwise. While this was going on Sally was busy contacting Len Gillings, a fellow member of their society, who owns a similar Tangye engine. Using information obtained from Len, Marc was able to “adapt” the magneto drive so that it produced a spark. He then managed to get the engine to start and run (this was all recorded in the visitors’ book).

The running had to be cut short for two reasons: the cooling system was drained, and the “adapted” magneto drive was causing damage.

Now that we had confirmed the direction of rotation we could concentrate our attention on increasing the magneto deflection. The first step was to make a longer catch plate with a slot to enable adjustment (see picture 2). This produced a maximum deflection at trip of 23 degrees and was successful in that we were now able to consistently start the engine, but running was ragged with plenty of pops and bangs, so back to the drawing board.

The poor running was attributed to faulty ignition timing, the longer catch plate in giving a bigger deflection was delaying the trip point. It was reasoned that this could be corrected by bending the longer catch plate in the direction in which it was pushing the lever of the magneto. A second new catch plate was made and we were then able to start and run the engine for short periods on petrol. We had to keep the running periods short as the cooling system was still drained.

The momentous day was July 1st, the cooling system had by now been refilled, the petrol container was full and about 1 litre of paraffin and been put in the main container. The engine was started and ran with a few misses but with no tendency to stop. After about 8 minutes the fuel cock was changed to paraffin. At first the engine started to slow so it was set back to petrol and then almost immediately after changed to paraffin and the engine continued to run for the next 15 minutes, until the fuel level dropped below the pick off point.

We thought we had won at last but this old engine still had a few tricks up its proverbial sleeve. The next time we started it up we just couldn’t keep it running long enough to change over the fuel supply. It kept slowing down and then picking up, stopping but restarting fairly readily. It was concluded that the most likely cause was that the fuel feed was partly obstructed. The whole of the petrol side of the fuel lines was removed, and lo and behold what looked like part of an old dead flower was found in the petrol inlet side of the fuel cock, together with a fair amount of very fine rust. The inside of the petrol container was cleaned using plenty of gravel and vigorous shaking; this released a considerable amount of rust and the inside of the container was now showing signs of bare metal.

All parts were restored on the engine, petrol was added but no paraffin. There were no apparent leaks and the engine was started and ran well. After about 2 minutes the engine was stopped. Paraffin was added and the engine restarted but stopped of its own accord after another few minutes with a massive fuel leak at the main fuel jet. On trying to tighten the gland nut it was found that the fuel jet had in fact sheared. We were able to find a local Engineering firm with the requisite skills to make us a replacement.

The new jet was fitted on September 16th 2002 and the engine started on petrol with minimum of effort. After some 5 minutes we were able to switch to paraffin and the engine then continued to run for another 40 minutes until the fuel was exhausted. It stopped a couple of times but was readily restarted on paraffin.

Work in restoring the pump continued until success was achieved in September 2010.

Credits and Acknowledgements

Ernest Samuel (Royston Essex)
For his advice and encouragement following our chance meeting at the 2001 Barleylands Show where he was rallying his Tangye Vertical Engine.

Mike Green (Independent Ignition Supplies, Appledore Bideford N.Devon)
For his work on the magneto, and subsequent patience and understanding.

Peter Thwaites (Tangye Engine Register, Mirfield W.Yorkshire)
For all the information that he supplied, including the fact that there are only four Tangye Vertical Engines remaining according to the official UK register.

Marc, Nick and Sally (Wychavon Oil Engine and Preservation Society)
For being in the right place at the right time.

Len Gillings (Bedworth Warwickshire)
For supporting Marc, Nick and Sally, and having his Sunday afternoon disturbed on April 21, also owner of the third Tangye Vertical Engine, which helps us date our engine to late 1920.

Gary Sefton (E & C Machining, Narberth, Pembrokeshire)
For his skills with the lathe and help in making the replacement fuel jet.

Back to top

Facts and figures

Back to top