Five- year conservation project gets underway at Tintern Abbey!

We are absolutely thrilled to announce that together with our friends at ArchaeoDomus we are supporting Cadw with an ambitious five year conservation programme at Tintern Abbey. The Abbey Church, which has stood for over 750 years, is in need of some help! The sandstone structure is beginning to decay and urgent conservation is needed. To enable rather complicated scaffolding to be designed we are supporting Cadw by investigating what impact the proposed scaffolding may have on the structure and below ground remains.

Together with Ross Cook (ArchaeoDomus) and David Robinson, the leading expert on Tintern Abbey and the Cistercians in Wales, we will be implementing a significant programme of excavation, building analysis, and historical research.  The results of the excavations will be made publicly available at the end of the project and will greatly enhance what’s known about Tintern and aid its future management and conservation.

Richard Lewis, Managing Director, Black Mountains Archaeology, said: “We are very excited to be supporting Cadw with this ambitious, landmark conservation programme for Tintern Abbey’s Great Church. This will be the first time that the Abbey Church will have been investigated in such detail since the Ministry/Office of Works carried out their landscaping and excavations nearly 100 years ago and offers an exciting opportunity for us to learn more about the magnificent building.

We are currently investigating a large area of the church and we are recording the monument using the latest modern digital archaeological recording techniques. We are using cutting edge photogrammetry and laser methods, together with bespoke digital software to record the investigations in 3D. This will provide a detailed, holistic and lasting record for the future.

Medieval floor tile

So far we have recovered many interesting artifacts from the late 13th century to the modern period. These include, fragments of rare medieval window glass, floor tiles and pottery,  and coins from Henry III (1216-72) to George III (1760-1820), through to the Victorian and Edwardian Periods.”

Medieval floor tile



Medieval window glass






Ross Cook, Buildings Archaeologist and Dendrochronologist at ArchaeoDomus, said “The nature and scope of the investigations at Tintern are unprecedented and will provide the most comprehensive record of a Scheduled Monument in Wales.

Silver shilling 1888

The programme of buildings recording will explore every inch of the monument and create a ‘living’ database of information relating to its pre- and post-conservation condition and historical development. This will continue to be added to in the future to support the management of this beautiful and important site. It’s a very exciting project to be a part of and great to work alongside Cadw and a wonderful team including acknowledged specialists in the history of the Cistercians and  masonry conservators.

King Henry III Short Cross Penny

The excavations have already helped improve our understanding of the construction and development of the Abbey Church in both the pre- and post-Reformation medieval world. Both myself and Black Mountains Archaeology look forward to sharing our findings in the coming months and years.”

You can learn more about this amazing project on Cadw’s website here and here, and below.

Pont ar Daf Car Park, Brecon Beacons – Archaeological Field Evaluation

Storey Arms Public House

Back in the summer of 2017 we were commissioned by National Trust Wales to carry out an archaeological field evaluation ahead of the proposed extension to the existing popular car park to inform on the nature and extent of any archaeological remains at Pont ar Daf Car Park in the Brecon Beacons National Park. The proposed development consists of the construction of a new extended car park of around 263 spaces in an area of former forestry and Post-medieval settlement. The car park construction is now well advanced.

There was a significant potential to encounter Prehistoric, Roman, Post-medieval and WWII defensive activity in the general vicinity. With particular emphasis on the presence of the Storey Arms WWII Anti-Invasion Stop Line (SAMBr337) and associated sites, the Beulah-Penydarren Roman Road (14900/14901/14902) and the old Storey Arms public house (115060), outbuildings and fieldscape.

Arguably the most significant discovery was that of the metalled surface found in Trench 6. It is possible that the poorly surviving metalled surface is the Beulah to Penydarren Roman road (RRX77 14900) suggested by Hogg and Houlder (1969), however, we did not recover any dating evidence to confirm this hypothesis. The road feature is very fragmented but appears to survive well, breaking the ground surface along the modern post and wire fence boundary. A 3D photogrammetric model of the potential road in section is shown above. However, much more investigative work will be required to ascertain the origins of the metalled surface, if indeed it is Roman at all.

Metalled road surface found in Trench 6

For further details please download the report here 108 Pont ar Daf Car Park, Brecon Beacons.

Royal Observation Corps Underground Monitoring Post, Llandrindod Wells

In September 2022, we were asked by PAR Homes Ltd and Asbri Planning Ltd to undertake an archaeological field evaluation in fields off Lakeside Avenue, on the southern end of Llandrindod Wells, Powys. In the northernmost field, we uncovered the demolished remains of a Cold War-era nuclear bunker, also known as a Royal Observation Corps (ROC) Underground Monitoring Post (UGMP). Within the vicinity of the bunker we also uncovered the foundations of an aircraft observation station, known as a Type B Orlit post. The presence of the UGMP was known prior to our evaluation, although whether or not it survived underground was not clear. The presence of the Orlit post was a lovely discovery and was not known prior to our investigations.

Aerial photograph of the Royal Observation Corps (ROC) Underground Monitoring Post (UGMP) during excavation

History of the Royal Observations Corps (ROC)

The ROC was established in 1925 as part of the British air defence system which, at this time, was rapidly trying to keep pace with, and potentially respond to, advances in aerial warfare technology brought on by the First World War (WWI). Between 1925 and 1941, the organisation was known simply as the ‘Observer Corps’ (OC). The main role of the OC was to identify hostile aircrafts flying over or within the vicinity of the British Isles. Prior to the establishment of the OC, during WWI, air defence was the responsibility of the Admiralty, to whom information on potential hostile aircrafts was sent via police reports. By 1920, after the end of WWI, British air defences had been rolled back, with most anti-aircraft hardware in the country being decommissioned. However, by 1924 Air Raid Precautions (ARP) guidelines were established, which were designed to inform on the reporting of, and countermeasures against, potential dangers to British civilians from air raids. In addition, ARP guidelines filtered through to urban and industrial planning, engendering a policy of continual aerial defence preparations during peacetime (Page 2016, 119). Within a year, the ARP had developed into a specified methodology for the OC. Prior to the outbreak of the Second World War (WWII), the ROC was plagued with technical and organisational issues, although these were ironed out in 1939 during a series of exercises intended as preparation for potential Luftwaffe raids. During the Battle of Britain in 1941, and the Blitz more generally, the role of the OC was integral to British air defence and aided greatly in allowing the RAF to fight off the Luftwaffe. OC personnel worked tirelessly in tracking enemy aircrafts and passing information on to RAF Fighter Command Groups and Sector Controls. In recognition of its importance to British defence, the OC was granted the title ‘Royal’ by King George VI in April 1941 (therefore becoming the ROC).

An ROC operative conducting air observations during WWII

By 1945, at the end of WWII in Europe, and the cessation of Luftwaffe raids on British territory, the ROC stood down. Yet during that same year, the ROC was once again put on active duty to deal with potential security threats posed by Britain’s involvement in the Cold War. Principally, it was the ongoing expansion of Soviet-style socialism into much of Central Europe, and the perceived threats that this posed to British interests, that led directly to the reformation of the ROC (Clarke 2005, 140). In 1954, the first thermonuclear weapon (or H bomb) was detonated by the USA during Operation Castle Bravo at the Bikini Atoll of the Marshall Islands. The radioactive fallout from this event was both devastating and (perhaps more importantly) extensive, meaning that in the wake of a thermonuclear explosion, large areas of land would be vulnerable to drifting radioactive particles. In 1955, the Strath Committee was established, which comprised a secret group of civil and military experts tasked to assess the potential impacts of thermonuclear weapons on Britain from an environmental, societal, defensive and economic perspective (Hughes 2010). The Strath Committee concluded that the detonation of only a limited number of thermonuclear weapons along the western seaboard of Britain would lead to the release of a devastating amount of nuclear fallout across the entire country (Clarke 2016, 206). As a response to this potential threat, the United Kingdom Warning and Monitoring Organisation (UKWMO) was established in 1957, which was responsible for, among other things, the monitoring of potential nuclear detonations and their ensuing fallouts. During that same year, the UKWMO set up a large network of UGMPs, numbering 1,518 in total, which were operated and maintained by the ROC. The placement of these posts formed a grid-like pattern across the country, with each post being situated in an isolated rural area. Moreover, each post was situated approximately seven to ten miles from its neighbour and were often positioned, for reasons of accessibility, on the periphery of playing grounds or arable fields (Bennett 2018, 215). With the introduction to Britain of UGMPs came the abandonment by the ROC of most of its aircraft observation duties (however, see references to Orlit posts below). Indeed, the principal focus of the ROC between 1957 and the end of the Cold War was nuclear monitoring.

Mushroom cloud resulting from the Castle Bravo thermonuclear explosion

History of Underground Monitoring Posts (UGMPs)

A detailed analysis of ROC UGMPs has previously been carried out by Robert Clarke (2016), where he identified the function and layout of these subterranean structures. A fantastic model of the interior of an ROC UGMP has been illustrated by Bob Marshall (see below). The work of Marshall serves as a useful, visual demonstration of life within, and the workings of, a typical UGMP. Clarke notes that each post was occupied by three or four ROC personnel, whose daily tasks primarily consisted of observation and submitting periodic reports to larger command and control bunkers (Clarke 2016, 76). In total, 31 of these larger bunkers were established across Britain, the construction of which was funded by the Home Office. Once the personnel within the larger bunkers received a report from those operating a given UGMP, the information within the report was then passed on to the military and Home Office (Wood 1992). Depending on the vagaries of local geology, each UGMP was buried to a depth of at least 4m (Clarke 2017, 242). In terms of composition, UGMPs were constructed using materials capable of withstanding the kinds of environments left in the wake of nuclear explosions. Predominantly, these materials included reinforced concrete, steel and tungsten (Clarke 2016, 89).

Digital Reconstruction Cutaway of a Royal Observer Corps Monitoring Post reproduced here with kind permission of Bob Marshall

In effect, each UGMP comprised a subterranean control room measuring approximately 5.8m x 2.6m in area and around 2.3m in height (Clarke 2016, 206). Accompanying the control room was a series of above-ground instruments, which recorded exterior radiation levels and monitored any potential detonations and the amount of energy they released. Initially, communications between UGMPs and elsewhere were often conducted via telegraph wires, which were highly vulnerable and susceptible to damage. In order to rectify this problem, radios were installed in the control rooms of UGMPs in the 1960s, which replaced telegraph communication. All radio equipment was situated within the master post, which comprised a small station positioned against the interior wall of the bunker. Within the master post was a VHF radio, carrier receiver, teletalk device and filter unit. Attached to one of the ventilation stacks above ground was an aerial mast, via which radio communication could be achieved. This mast could be extended into the air via an air pump located within the control room. Also situated above ground was a Ground Zero Indicator (GZI) – an instrument designed to locate the hypocentre of a nuclear explosion. These instruments comprised circular metal boxes with four horizontally mounted pinhole cameras inside, within which sheets of photosensitive paper were placed. If a nuclear blast was set off within or near Britain, the resulting bright flash would burn a mark on the photosensitive paper. The position of this mark on the paper would inform on the height of the nuclear blast, with the position of the blast being determined by triangulating data from GZIs situated at neighbouring UGMPs. The GZI was situated on top of a brick or concrete plinth, with a circular mount fixing it in place. This mount was attached directly to the plinth. Sometimes, the plinth was positioned above ventilation duct of the main entrance stack, as was the case with the Llandrindod Wells UGMP (see below).

In 1991, the Soviet Union was dissolved, and the Cold War ended. This, in turn, led to a diminishment in nuclear hostilities and, therefore, a reduction in nuclear threats to Britain. Despite the easing of tensions that followed the ending of the Cold War, the ROC was kept in active duty throughout the early–mid-1990s, albeit in a much-reduced capacity. The remaining personnel attached to the ROC were at this point, however, confined to maritime and not terrestrial duties. By the summer of 1990, in the face of radical reforms to the Soviet Union and the Warsaw Pact proposed by Gorbachev, the British government commenced the Options for Change programme, which aimed at reducing military spending. The implementation of this programme ultimately led to the disbandment of all remaining UGMPs in Britain.

Results of the Archaeological Field Evaluation

During our archaeological field evaluation, a trench was excavated across the proposed position of the Llandrindod Wells UGMP in order to determine its presence and state of preservation. As we found out, this was easier said than done! The small size of the bunker, combined with the masses of demolition material spread across the field, made its discovery more difficult than we anticipated. We did have in our possession some photographs of the bunker’s entrance stack before its demolition (see below) which, in terms of their ability to accurately determine the position of the bunker, were limited. After a few attempts, however, we found it. Although no part of the UGMP was found to survive, its broad outline could nonetheless be observed. Unfortunately, due to its previous demolition, the internal arrangement of the bunker was not ascertained, and neither was its structural form. However, the finds collected from within the demolition material of the bunker were able to shed light on some of the activities conducted within and around its confines. The archaeological remains encountered during our field evaluation were recorded photogrammetrically, which allowed them to be preserved as digital 3D models (see below).

Photogrammetric orthomosaic of trench excavated through demolished ROC UGMP (left) and remains of the Type B Orlit post (right)

Photogrammetric orthographic image of east-facing of section of trench excavated during field evaluation through demolished ROC UGMP

The most notable finds retrieved from the demolition material of the UGMP included a Ground Zero Indicator (GZI) mount, a master post carrier receiver, and a master post telecoms filter unit. The GZI mount comprised an upper circular fixture attached to a triangular base via a circular neck (see below). The mount was cast in one piece (in cast iron) and was painted a dark green colour. The GZI (discussed above) was an essential component of the UGMP as it could measure the position and size of any potential nuclear explosion within the UGMP’s vicinity. Amazingly, we had in our possession a photograph of the GZI mount when it was in situ (see below). In this photograph, the mount is shown as being attached to the concrete forming part of the top of the ventilation duct attached to the entrance stack.

Photograph of the Llandrindod Wells ROC UGMP prior to its demolition in the 1990s, showing the entrance stack as well as the GZI mount attached to the ventilation shaft (in red)

The GZI mount collected during the archaeological field evaluation, with part of the concrete once comprising the ventilation shaft of the UGMP’s entrance stack still fixed to its base

The master post carrier receiver comprised a grey plastic box with a speaker grill attached to the front. The find incorporated a front cover, on which the speaker grill was located, which allowed it to be opened for inspection/repair. On this cover was a button and a dial, situated below the speak (microphone) grill. The button, situated on the left-hand side, was for testing the radio and speaker, while the dial, situated on the right-hand side, controlled the volume. A serial number was noted on the right-hand edge of the find, which read ‘EMT 82/21’. In terms of function, the carrier receiver was originally attached to the master post, which comprised a set of equipment mounted to the wall of the bunker that collected readings and messages from derived from the local cluster of UGMPs to which the bunker belonged. The carrier receiver allowed the personnel manning the UGMP to receive the National Attack Warning Red (implying aerial attack), the Fallout Warning (implying a nuclear strike and fallout) or All Clear (implying no immediate danger. Two models of the carrier receiver were produced for use within UGMPs, known as the 1st and 2nd generation. The 1stgeneration model was notably small and comprised a speaker with no buttons or dials. The 2nd generation model was notably larger and comprised a speaker with a test button and volume dial. The carrier receiver recovered during from the demolition was a 2nd generation model.

The master post carrier receiver recovered during the archaeological field evaluation

The master post telecoms filter unit comprised a small black box, rectangular in profile, with a detachable cover on the front held in place with screws. The find had bevelled (or rounded) edges, while on the interior was a circuit board a wires. On the front of the find, in white print, were the words ‘TELECOM’, ‘FILTER UNIT’ and the model number ‘WB1410’. In terms of function, the filter unit selectively sorted incoming telecommunications signals through a specified (and desired) range, while suppressing others. More specifically, the WB 1410 filter separated the carrier from the speech and power received by the teletalk, which represented the main communication tool within the UGMP bunker.

The master post telecoms filter unit recovered during the archaeological field evaluation

Finally, one of the more unexpected discoveries of our archaeological field evaluation were the remains of the Type B Orlit post. As mentioned above, the presence of this feature was not known prior to our archaeological field evaluation. The remains of this feature were uncovered to the east of our trench, during initial attempts to locate the approximate position of the UGMP. Our discovery of the Orlit post, therefore, occurred almost by chance! The surviving remains of the Orlit post comprised four concrete foundation pads, which together formed a broadly square shaped footprint. The southeastern foundation pad had remnants of brickworks attached to it, indicating that it once held a column above it. Not one the photographs we had in our possession of the UGMP showed the existence of an Orlit post within the area. This indicated that the post was likely demolished during the use of the UGMP. Clarke, in his research of Orlit posts, describes it as “an overground observation point for aircraft reporting….constructed using pre-cast concrete panels manufactured to a design by Messrs Orlit, Bedfordshire” (2016, 14). ). It is clear that Orlit posts ranged quite significantly in form and size, with two types being defined by Clarke. The first, Type A, was constructed over a concrete base at ground level. The second, Type B, was constructed on four precast concrete legs (Clarke 2016, 162). It appears that the feature we discovered can be identified as a Type B Orlit post. From this finding, we now know that within the Llandrindod Wells UGMP the above-ground aircraft reporting was, for a time at least, being conducted in conjunction with underground nuclear monitoring. In form, the Type B Orlit post was effectively a raised concrete box accessed via a wooden door, within which communications board, telephone and various other equipment were located (see below). On the right-hand side of the entrance, a sliding door was often incorporated to provide access to a small observation platform, which was open to the elements albeit for a moveable, corrugated tin cover. On top of the Orlit post, an instrument was mounted, which plotted the bearing and altitude of unidentified or potentially hostile aircrafts. Despite the importance of the Orlit post to the UGMP and the intricacy of the activities conducted within, the structure was markedly rudimentary, and when viewed from the exterior it had the appearance of a simple store shed.

Diagrammatic section and plan of an ROC Type B Orlit post (after Wood 1992)

Selected Bibliography

Bennett, L. 2018. ‘Cold War Ruralism: Civil Defense Planning, Country Ways, and the Founding of the UK’s Royal Observer Corps’ Fallout Monitoring Posts’. Journal of Planning History 17 (3), pp. 205–25.

Clarke, R. 2005. Four Minute Warning: Britain’s Cold War. Stroud: Tempus.

Clarke, R. 2016. Landscape, Memory and Secrecy: The Cold War Archaeology of the Royal Observer Corps. Unpublished PhD Thesis: University of Exeter.

Clarke, R. 2017. ‘The Secret Taskscape: Implications for the Study of Cold War Activities’. In U. Rajala and P. Mills (eds), Forms of Dwelling: 20 Years of Taskscapes in Archaeology, pp. 236–52. Oxford: Oxbow.

Hughes, J. 2010. ‘The Strath Report: Britain Confronts the H-Bomb, 1954–1955’. History and Technology 19 (3), pp. 257–75.

Page, A. 2016. ‘Planning Permanent Air Raid Precautions: Architecture, Air War and the Changing Perceptions of British Cities in the Late 1930s’. Urban History 43, pp. 117–34.

Wood, D. 1992. Attack Warning Red: The Royal Observer Corps and the Defence of Britain 1925 to 1992. Portsmouth: Carmichael and Sweet.


Melin Mynach, Gorseinon

We were recently asked by Swansea Council and Mann Williams Civil Engineers to undertake a 3D photogrammetric survey of the Melin Mynach Scheduled Monument (SMGm501), a former mill site in Gorseinon, to inform on conservation management plans.

The site broadly consists of the remains of a medieval and Post-medieval gristmill, an early Post-medieval paper mill and later woollen mill. Cadw note that “…water power has been used at this location for a number of industrial purposes for a long period [of time]. The earliest mill is thought to be of monastic origin, possibly built after Neath Abbey took over the estate of Cwrt-y-carnau in 1150, and was probably a corn mill. The first documentary reference to a mill was in 1578. In 1772 it was converted for paper making, and was one of the first of its kind in Wales, in use for over a century. It was returned to use as a corn mill in the 1830’s, but in 1866 William Lewis converted the mill to woollen manufacture, enlarging it substantially in 1874. From 1888 the site was turned to chemical and tinplate manufacturing, and the mill itself became disused. The main surviving features are the leat, pond, paper mill, woollen mill, two wheel pits, dye-houses and the mill owner’s house” (Cadw).

The survey was undertaken with a high resolution camera equipped UAV (drone) and DSLR camera. The 3D photogrammetric model was produced using proprietary photogrammetry software and aligned using known ground control points (GCPs), which were tied into the Ordnance Survey National Grid and Datum using an RTK GNSS/Glonass (GPS) Receiver. The resulting Ground Sampling Distance (GSD) achieved 0.83cm/pixel. The 3D model produced a dense point cloud of over 171 million points and a high face count mesh (>34m), with a mean RMS error of 6mm. You can view a copy of the 3D photogrammetric model below that has been reduced in size/detail. The finished 3D model was so large (and detailed) we had to reduce it by around 70% to allow us to share online. 

We then produced series of high resolution measured orthographic plans and elevations to inform on the conservation efforts.

Orthographic NE and SW Facing Photogrammetric Wheel Pit Elevations

The Medieval and later Post-medieval Development of the Mill

The historical and archaeological background of Melin Mynach has previously been outlined in great detail by Martin Lawler (1990). So we’ve summarised some of his work below and augmented this where necessary with more recent archaeological developments to provide a little historical background to Melin Mynach.

Medieval and Post-medieval Gristmill

The medieval gristmill, also known as the Monk’s Mill, is understood to be of monastic origin. The site of the mill formed part of a grange attached to the east of Neath Abbey, which was established in around 1150. This monastic grange was known as Cwrt-y-Carnau. At the beginning of the 19th century, a sizeable farmhouse (NPRN18500) was erected on the site of this former grange (Roberts 2014, 228). In terms of extent, the grange spanned an area between Loughor Estuary in the west to Afon Lliw in the east. It has been suggested by Williams (2001, 308) that the grange centred on NGR SS 593 988, which denotes a point on the eastern edge of Gorseinon. In total, the Cwrt-y-Carnau Grange formed part of a collection of seventeen estates over which Neath Abbey had control. The earliest reference to the gristmill at Cwrt-y-Carnau belongs to the sixteenth century and is recorded as occupying an area towards the western banks of Afon Lliw. As this reference was made several centuries after the construction and use of the Cwrt-y-Carnau mill, the process of definitively identifying the Melin Mynach site as its place of origin proves difficult. However, the Melin Mynach site is situated near the western banks of Afon Lliw and is the only site of this kind in the area. It may therefore be approximated that the gristmill at Melin Mynach did indeed form part of the Cwrt-y-Carnau Grange. In terms of its position, the mill would have been isolated from the chapel and grange centre, which were situated further west towards the Loughor Estuary. The chapel to which the mill originally belonged is known as St Michael’s, which today comprises a Scheduled Monument (SMGm363). Yet this was not an unusual situation for the time, as some other monastic granges in Wales were similarly arranged, such as those belonging to Strata Florida Abbey in Ceredigion and Whitland Abbey in Carmarthenshire.

The surviving features associated with the medieval gristmill include a small section of mill race, the mill pond and the earthen platform on which the mill itself was constructed. The mill race was one of the longest of its kind in Wales, measuring over 2.8km in length. The construction of this mill race, moreover, involved the highest engineering expertise that the twelfth and thirteenth centuries had to offer. The head race section diverted water from Afon Lliw towards the western edge of the Lliw Valley to a height of approximately 4m above the river. The tail race section then returned the water to Afon Lliw at a point near King’s Bridge. In terms of structure, little of the original medieval gristmill survives as it was reconstructed in the eighteenth century. Akin to other industrial sites belonging to monastic granges, the gristmill would likely have been managed by lay brethren. Following the Act of Supremacy 1534 and the resulting dissolution of the monasteries in England and Wales, the gristmill, along with the entirety of the Cwrt-y-Carnau Grange, was transferred to private ownership. As a result, the mill at this time formed part of a manorial demesne until as late as the nineteenth century. It was during this time that the mill was reconstructed and enlarged.

Post-medieval Paper Mill

The conversion of the former medieval mill into a combined grist and paper mill was commissioned during the early eighteenth century by Thomas Selman, the then proprietor of the mill. At this time, Melin Mynach represented one of the three earliest paper manufacturing sites in Wales. An earlier paper mill situated on the edge of Kidwelly in Carmarthenshire was previously owned by the Selman family. Yet for reasons unknown, the Selmans decided to transfer their paper manufacturing business from Kidwelly to Melin Mynach by 1729. Structurally, the new paper mill at Melin Mynach was composed of sandstone rubble walling.

The manufacture of paper at the mill involved the recycling of old textiles in the form of, for example, rags, cloth and cordage. These textiles were first cleaned and separated into individual fibres, in an operation similar to carding.

Grading Rags – Diderot Encyclopedie 1767

The separated fibres were then chopped and boiled before being transferred to the pulp mill which, it is presumed, would have been located near the water wheel.

Stamping – Diderot Encyclopedie 1767

Within the pulp mill, the fibres would have been pulverised via a series of timber stampers, which were operated via cams attached to the shaft of the water wheel. Moreover, the fibres were placed on a series of stone or timber mortars, meaning they were pulverised between the base of the stamper and the top of the mortar. These mortars were graded, with each grade facilitating a slightly finer texture of fibre.

Machine stamping – Diderot Encyclopedie 1767

The fibres were, therefore, transferred between each grade consecutively, before a desired texture was obtained. The precise nature of this mortar grading and the ways in which it operated in unknown. This process took between twelve and 36 hours to complete, depending on the texture of the type of textile being worked. Once this process was completed, the resulting pulp from the mortars was transfered into a large vat, which was attached to a charcoal heated stove.

Moulding Vat – Diderot Encyclopedie 1767

Once within this vat, the pulp was continually agitated through stirring. A fine sieve was then used in order to mould the portions of the pulp into individual sheets. Each sheet was placed onto a corresponding ‘felt’ (composed of woollen cloth) and a stack of alternating sheets and felts was formed. This stack, or ‘post’, generally consisted of 144 pairs of sheets and felts. Once these posts were formed they were compressed in a screw press, which reduced the thickness of the post from around 600mm to approximately 150mm. The individual sheets could then be separated from their corresponding felts to be hung up and dried. These individual sheets, at this stage of the manufacturing process, took a form that could be described as paper.

Drying -Diderot Encyclopedie 1767


Post-medieval Woollen Mill

By the 1860s, the paper mill at Melin Mynach was out of use and was purchased by William Lewis, a woollen manufacturer from Penllergaer to the east of Gorseinon. Lewis had converted the old paper mill into a woollen mill by 1875, a process which involved the construction of a new mill building.

1874 date stone for the new woollen building

The old paper mill was, however, retained rather than demolished and may have been used by Lewis as a weaving shed. The new mill building was three storeys in height with a large waterwheel on its northern end. This waterwheel was reconstructed in the 1990s. In terms of structure, the woollen mill was more robust than the previous paper mill and was constructed using roughly dressed sandstone blocks. Internally, the woollen mill was compartmentalised into four bays and the windows incorporated into its longitudinal east wall were markedly large in size, spanning both the ground and first floors of the building.

Melin Mynach Reconstruction Drawing. How it may have appeared in the late 19th Century. Reproduced with kind permission of the Glamorgan-Gwent Archaeological Trust.

The precise operations being conducted within the woollen mill are difficult to determine, although it is likely, based on information from Lewis’ account books, that the interior was supplied with carding and scribbling machines. These same account books also indicate that William Lewis was supplying woollen flannel manufactured at the mill to the local markets of Swansea, as well as cloth for shirts and underclothing.  Attached to the southern end of the new mill building was the dyehouse, with passage between the two buildings being facilitated by a pair of parallel doorways. Beyond the dyehouse there exists evidence that the woollen mill witnessed further extension during the Lewis family’s proprietorship. These building extensions are presumed to have bee incorporated sometime after 1879. One of these extensions takes the form of a T-shaped building, single storey in height, with narrow sandstone footings. Another extension is situated approximately 13m of here and takes the form of another narrow single storeyed building. The precise function of these buildings has not yet been determined.

Orthomosaic Plan of Melin Mynach

Selected Bibliography

Lawler, M, 1990, Melin Mynach, Gorseinon: Archaeology, History and Future Prospects, GGAT Report No 90/01.

Roberts, R, 2014, Cistercian Granges in Glamorgan and Gwent, GGAT Report No 2014/023.

Williams, DH, 2001, The Welsh Cistercians, Bodmin, MPG Books.

The Encyclopaedia of Diderot & d’Alembert, 


Former Tyla Cartshed, Gilwern Hill

We recently carried out a survey on a small stone outbuilding adjacent to Lilac Cottage on Gilwern Hill. You can view the 3D photogrammetric model here and download the full building record here ‘213 Lilac Cottage Stable Building Survey Report’.
The outbuilding is recorded by the RCAHMW (NPRN270200) as Tyla Cottage Stable. Although the partial archway in NE facing elevation would suggest that it was in fact a much larger cart shed, with later phases reducing the size of the building and the infilling of the arch. They note anecdotal evidence to suggest that Lilac Cottage is said to have been the former Quarryman’s Arms public house (NPRN270203). The Ordnance Survey 1st Edition map (1888) records the cottages here as ‘Tyla’, perhaps referring to a hamlet name for the quarry workers houses.

Ordnance Survey 1st Edition historic map (1888)

The Tyla settlement is named for the limestone quarry to the south and west. The Tyla quarries were the first limestone quarries opened up for the fledgling ironworks at Blaenavon after 1787. The limestone was used as a flux to stick slag together in the furnaces as well as mortar to build. The Hanburys of Pontypool had been extracting ore from the mountains around Blaenavon since Tudor times using scouring methods on the mountain top, and this was continued by the Blaenavon partners after 1787 to feed the ironworks. In 1787 Thomas Hill and Benjamin Pratt from Stourbridge successfully negotiated a 21 year lease of some 12,000 acres, formally leased by the Hanburys, from Lord Abergavenny. They were then joined by Thomas Hopkins, a Welshman who had managed forges at Rugeley, Staffordshire. Together they constructed three purpose-built furnaces and cast houses and were producing pig iron by 1790, some 3,600 tons by 1792. By 1806 furnaces 5 and 6 had been built and then later Hill’s Tramroad and Garnddyrys Forge were operational by 1817.

Garnddyrys Forge and Hill’s Tramroad in the early 19th century. Reconstruction by Michael Blackmore.

The additional furnaces at Blaenavon and the growing lime trade on the Monmouthshire and Brecon Canal created further demand for limestone. With large limekilns built on the canal at Llangattock, Gilwern, Llanfoist and Goytre. Tyla quarry had been opened up after 1787 and initially limestone was hauled by horse on tracks over the mountains to Blaenavon. By around 1800 a tramroad had been constructed from the quarry to Pen-ffordd-goch (Keepers Pond) and down to the ironworks at Blaenavon. This route was abandoned when Hill’s Tramroad and the Pwll Du Tunnel (and quarries) was operation in around 1817.
The Reverend William Coxe during his Historical Tour of Monmouthshire (1801) marvelled at a new ‘rail road’ being built from the Tyla Quarries and his description is a useful insight to industrial activity at Tyla:
In the vicinity of Blaenavon we observed the process of making a rail road, so called because it is formed by a kind of frame with iron rails, or bars, laid length ways, and fastened or cramped by means of cross bars. The ground being excavated, about fix feet in breadth, and two in depth, is strewed over with broken pieces of stone, and the frame laid down, it is composed of rails, sleepers, or cross bars, and under sleepers. The rail is a bar of cast iron, four feet in length, three inches thick, and one and a half broad; its extremities are respectively concave and convex, or in other words are morticed and tenanted into each other, and fastened at the ends by two wooden pegs to a cross bar called the sleeper. This sleeper was originally of iron, but experience having shewn that iron was liable to snap or bend, it is now made of wood, which is considerably cheaper, and requires less repair. Under each extremity of the sleeper is a square piece of wood, called the under sleeper, to which it Is attached by a peg. The frame being thus laid down and filled with stones, gravel, and earth, the iron rails form a ridge above the surface, over which the wheels of the cars glide by means of iron grooved rims three inches and a half broad.”

Copperworks Discovery Project 2021

We are excited to announce a new community archaeology project at the world-renowned Hafod-Morfa copperworks, made possible by a grant from the National Lottery Heritage Fund. We are on the hunt for volunteers! No experience or equipment is necessary, just come along and enjoy digging a piece of Swansea history. To register your interest please email Abbi at

If you missed the virtual launch event on Tuesday 26th October 2021 then you can catch up with a recording of the event on our YouTube channel here.

Don’t forget, the deadline for registering your interest for the archaeological dig is 10am Friday 29th October 2021.

#HeritageFund #NationalLottery #heritageforeveryone #communityarchaeology #industrialarchaeology Swansea Council

Rydym yn gyffroes i gyhoeddi prosiect archeoleg gymunedol newydd yng ngweithiau copr byd-enwog Hafod-Morfa, yn bosibl trwy grant gan The National Lottery Heritage Fund. Rydym yn hela am gwirfoddolwyr! Nid oes angen profiad nac offer, jest dewch draw i fwynhau cloddio darn o hanes Abertawe. I gofrestru’ch diddordeb e-bostiwch Abbi ar

#HeritageFund #NationalLottery #treftadaethibawb #archeoleggymunedol #archeolegddiwydiannol Cyngor Abertawe


Brief history of the Hafod-Morfa Copperworks

View of the Hafod Copperworks – from George Grant Francis (1881), The Smelting of Copper in the Swansea District, from the time of Elizabeth to the present day

During the mid-19th Century the Hafod-Morfa Copperworks employed over 1000 people. Situated in the Lower Swansea Valley, this area at that time accounted for 90% of the world’s copper production. This was made possible by a abundant supply of coal in the Swansea Valley, brought down via the late 18th century canals, and the excellent facilities for shipping, which allowed the import of copper ore from Cornwall, North Wales, South America, South Africa and Australia.

Copperworks in Lower Swansea Valley 1865 Le Tour du Monde
Copperworks in the Lower Swansea Valley 1865 – Le Tour Du Monde

The Hafod Copperworks was established in 1808-9 by the Cornish entrepreneur John Vivian. In 1828 a Cornish firm, Williams, Foster & Co., opened the Morfa works on adjacent land. The works was initially a rolling plant for making bars and plates from copper ingots brought from the nearby Rose Works but smelting is believed to have started by 1835. Both the Hafod and Morfa works amalgamated in 1924 and was subsequently operated by Yorkshire Imperial Metals until it closed in 1980, when it was the last operating copperworks in Swansea.

At least fifteen significant structures, in varying degrees of condition, survive across the Hafod-Morfa Copperworks site. These include the rolling mill (LB 16878) now used as the museum stores, the laboratory building (LB 11690) and the former Morfa Powerhouse and later Yorkshire Imperial Metals canteen (LB 11691). The Hafod Limekiln (11694), Copper Slag Abutment, Pier and Canal Boundary Walls (LB 11692 and 11693). The Vivian Engine House (LB 11695), the Chimney (LB 11696) west of the Vivian Engine House and the Boundary wall for the Hafod Copperworks Canal Docks (LB16881). Finally, the in-situ Musgrave Engine and Rolls (SAMGm483) in the Musgrave Engine House and Chimney (LB 11697).

You can check out some our previous investigations at the Hafod-Morfa Copperworks here!

3D Photogrammetric Survey of Alina’s Chapel, Oystermouth Castle

We were recently asked by Taliesin Conservation to undertake a high resolution, metrically accurate 3D photogrammetric survey of Alina’s Chapel, Oystermouth Castle (SAMGm007) ahead of conservation work. The survey was undertaken with a camera equipped UAV (drone) and tied into the Ordnance Survey National Grid and Datum using an RTK GN55/ Glonass (GPS) Receiver.

Oystermouth Castle Plan (c) RCAHMW

Oystermouth Castle is located on a high carboniferous knoll overlooking the town and Swansea Bay. The castle dates to the 12th century, probably built by William de Londres after Henry de Beaumont of Warwick’s conquest of Gower in 1107. The RCAHMW (2000, 249) note at least six phases of masonry development at Oystermouth (RCAHMW 2000, 245-272).

Aerial view of Oystermouth Castle

The castle most likely started as ringwork before being replaced by stone keep by Maurice de Londres (1138-41) (Period I).

John de Braose (1220-32) constructed the central block, a two-storeyed structure over a vaulted cellar set against the north wall of the keep (Period II).

Period III may post-date the Welsh attack on Gower in 1256. The northwest block and west range being added.

William de Braose II (1241-90) likely built the gatehouse and curtain walls, finished by the time Edward I visited the castle in 1284 (Period IV).

The chapel block (Period V) consists of the finest masonry in the castle and likely dates to the 14thcentury. Before construction it was necessary to demolish part of east curtain wall. The chapel consisted of a basement, used as a kitchen, first floor apartment and second floor chapel lighted by five decorated windows. The largest eastern window divided by two chamfered mullions and cusped and interlaced tracery (RCAHMW 2000, 245-272). The roof of the chapel forming a crenelated parapet. The chapel may have been built by Alina de Braose (1327-31), daughter of William de Braose III and widow of John de Mowbray.

Period VI consisted of the construction of several ranges set against the southwest and east curtain walls. The east range being a kitchen of some kind with excavations in 2009 (Sherman 2012, 6) recovering stone and slate roof tiles and a shell midden containing large quantities of oyster, periwinkle, muscle, limpet whelk and pod razor shell all potentially dating to the 13th/14th centuries.

The survey generated a dense point cloud over 97 million points with a mean RMS error of 0.013m. Six GCPs were used and surveyed with an EMLID Reach GN55/ Glonass (GPS) Receiver and data logger with a sub-20mm error margin to OSGB36 (National Grid). The Ground Sampling Distance (GSD) achieved was a great 0.84cm/pixel.

You can view the 3D photogrammetric survey of Alina’s Chapel here. A high resolution still from the 3D model is below.

3D Photogrammetric Orthographic of Alina’s Chapel, Oystermouth Castle

With thanks to:

We are very grateful to Taliesin Conservation, Swansea Council and Cadw for help and support during the project.

Newman, J, 2001, The Buildings of Wales, Glamorgan. Second Edition. University of Wales Press.

RCAHMW, 2000, An Inventory of the Ancient Monuments in Glamorgan Vol III – Part lb Medieval Secular Monuments The Later Castle From 1217 to the Present. Aberystwyth.

Sherman, A, 2012, Recent archaeological works at Oystermouth Castle: Archaeological evaluation, community excavation, watching brief and window recording. GGAT Report 2012/071.

Tintern Abbey 3D Photogrammetric Survey


We are pleased to be able to share some of our recent work in Monmouthshire. We were asked by Cadw to undertake a photogrammetric survey at Tintern Abbey (SAMMm102). The aim of the survey was to produce accurate, measured 3D photogrammetric modelling of the east wall of the Chapter House and east facing elevation of the Monk’s Dortor/Day Room. The survey areas were situated adjacent (east) to the mid-12th century cloister and forms part of challenging restoration work being carried out by IBEX Technical Access LTD. We also had time to take a quick drone video during lunch, you can view the video here.

Aerial view of Tintern Abbey

The Monk’s Day Room survey created a dense point cloud of over 18 million points with a mean RMS error of 0.8cm. The Ground Sampling Distance (GSD) achieved was a great 0.43cm/pixel. The 3D model of the Monk’s Day Room can be viewed here.

Monk’s Day Room 3D photogrammetric orthofacade

The Chapter House survey created a dense point cloud of over 36 million points with a mean RMS error of 0.8cm. The Ground Sampling Distance (GSD) achieved was a great 0.38cm/pixel. The 3D model of the Chapter House can be viewed here.

Chapter House 3D photogrammetric orthofacade

Aerial view of Tintern Abbey

The excavation of a 19th century cottage in Pontsticill

We recently carried out an excavation in the lovely village of Pontsticill. We were initially asked to undertake an archaeological watching brief during groundworks for a new build house. Historic maps showed the presence of several houses dating from the 19th century in the area of the proposed new house, which was just a garden at the time of the investigations. During the course of the archaeological watching brief we revealed the remains of a 19th century cottage known locally as Bryn Teg. The remains of the house were fully excavated and recorded in 3D using photogrammetry.

1842 Tithe Map

The 1842 Tithe Map (Plan of the Parish of Vaynor in the County of Brecon) records the area of the new build house occupied by a pair of small rectangular houses, now demolished, and associated gardens. These buildings were situated within Land Parcel 582, which is recorded in the 1842 apportionment as being a meadow belonging to William Jenkins and Phillip Watkins. The northernmost house was situated in an area now occupied by the Dolgaer Houses, which first appears on the 1885 Ordnance Survey (OS) Map (Brecknockshire XLVI.NW). The southern house also appears to have been replaced at this time by the new Bryn Teg cottage, located in the north-west corner of Land Parcel 582.

1885 OS Historic Map

We were thrilled to have been given a photograph by a local resident that shows the 19th century cottage Bryn Teg in the years prior to its demolition. A precise date for this photograph is unclear but the small field situated to the east of Dolgaer Houses is being used for cultivation, which marries well with another aerial photograph dated to 1972. Therefore, the demolition of the cottage must have happed around this time.

Aerial photograph of Bryn Teg cottage before demolition

The aerial photograph clearly shows a three storey house with chimneys on each gable end wall. Two local residents of Pontsticill, Graham Williams and Mike Burns kindly shared their memories of Bryn Teg with us. Graham remembered that he used to deliver newspapers to the house as a child.  Graham and Mike also remembered that during the mid-20th century the house was being rented by a roofer named Jack Andrews and his family, who came to Pontsticill from East London.

Orthographic oblique photogrammetric 3D survey of Bryn Teg

The archaeological investigations consisted of the full excavation and preservation by record of the remains of Bryn Teg cottage, formally a three storey house. The 19th century house was recorded with a blend of RTK GPS survey, 3D photogrammetry (derived from both aerial and terrestrial cameras) and a descriptive account and phasing of the visible internal and external elevations, as well as its flooring and associated internal features.

Bryn Teg cottage phase plan

The remains of the 19th century house included the front (west) retaining wall; the southern gable end wall; the northern gable end wall, which included an in situ chimney breast, spiral staircase and cast iron range (complete with a kettle!!); the rear (east) wall; and a pair of internal partition walls. The north-facing elevation of an exterior retaining wall, which supported a rear yard to the south of the house, was also recorded.

South facing gable wall with in-situ cast iron range and chimney staircase

The full 3D photogrammetric model of the 19th century house can be viewed here: You can also download the full archaeological watching brief report here 206 Land to Rear of Dolgaer, Pontsticill WB Report.

The 3D photogrammetric model produced a dense point cloud of over 31 million points with a mean RMS error of 0.7cm. The Ground Sampling Distance (GSD) achieved was a great 0.15cm/pixel. High resolution orthographic renders (orthoplanes and orthomosaics) were also produced (see plan and elevation above).


We are very grateful for all of the help we received from the residents of Pontsticill during the course of the investigations.

We are especially grateful to Alison and Steve Cox for being so supportive throughout the project. 

Thank You!