Geotechnical and Geological Engineering, 1996, 14, 21--40

TECHNICAL NOTE

An historical review of landslide research in the South Wales coalfield A. A L M A S M O U M , * S.P. B E N T L E Y t and H.J. SIDDLE$ *Department of Civil Engineering, University of Umm AI Qura, Makkah, Saudi Arabia, ~School of Engineering, University of Wales Cardzff,, PO Box 925, Newport Road, Cardiff CF2 lYE Wales, UK and $Str William Halcrow and Partners, 31-33 Newport Road, Cardiff CF2 lAB, Wales, UK

Received 1 3 D e c e m b e r 1994 Accepted 1 2 D e c e m b e r 1995

Summary The areal density of landslides in the coalfield of South Wales is one of the highest in the UK. During the past 100 years landsliding has had considerable impact, causing structural damage and loss of life. Most of the landslides were initiated under periglacial conditions but many became reactivated due to the activities of humans, particularly during the late 19th century when widespread urban and industrial development commenced in the Welsh valleys. A number of the area's larger landslides are first-time slides which occurred during the past 100 years. This paper sets out to chart the history of landslide research in the coalfield, which began through work by mining engineers. Keywords: Slope stability, landslide, mining, subsidence, history, research, coal measures.

Introduction The urban development of South Wales is intrinsically linked to the Industrial Revolution. Until the mid-1700s the economy of the South Wales valleys was based entirely on farming. At about this time the iron industry commenced using the local resources of ironstone, limestone, wood and water. D e m a n d for iron continued to rise due to ship building and the establishment of railway infrastructures both in Britain and abroad; by 1857 production peaked at nearly 1 000 000 tons per annum. With the decline in the iron industry, coal production became more important to the regional economy. By the 1850s the 'Coal Rush' had begun in earnest and again thousands 0960-3182 9 1996 Chapman & Hall

22

A l M a s m o u m et al.

of migrant workers were attracted to South Wales. Settlements began to appear around pitheads and steadily grew. In 1870, 65 % of coal exported from Britain was from the South Wales coalfield. Peak production of 58 million tons was achieved in 1913. With the fall in coal production from 1913, the influx of migrant workers declined and, with the depression of 1921-1939, the trend was reversed completely with 400 000 people leaving South Wales during that period. Population trends for the Rhondda Valleys are illustrated in Fig. 1. Today the urban environment of South Wales is a legacy of its industrial past. Most of the buildings date back to the latter part of the 19th century, when urban expansion was at its height.

Geology and mapping The South Wales coalfield is one of the major coalfields in the UK (Fig. 2) covering an oval-shaped area of some 2700km 2. The geological structure is that of an east-west trending, asymmetric downfold in which up to 2500 m of Coal Measures are preserved. The Coal Measures are divided by marine bands into Lower, Middle and Upper units (Woodland et al., 1957), all of which are characterized by cyclic sedimentation. The

(xlO00) 60

SO

Ix 100) - 200 COAL

A

I

PROOUCTION~if~/~

~

150

PULATION

40

COAL

PRODUCTION (TONS) 30

(SOUTHWALES)

POPULATION ( RHONDOA) 100

2O

10

/ I

/

SO

/ I

I

I

I

I

I

I

I

I

I

18'70 1890 1910 1930 1950 19'70 1880 1900 1920 1940 1960 Fig. 1. South Wales coal production 1874-1945 and Rhondda Valleys population trend

Hdlstone Gr=f

CgrbonfferousL=mestone

~

~

.

.

~ .

.

"

"

.......... ::::::::::::::::::::::::::::::::::::

:::::::::::::::::::::::::::::::::::::::::::::::

i

~'~!!

::

~'i::b 6

:

~~s'~:)4:::Jii!i5i?i!+[iiiiiis

,/

7"

:::: ::iiiii~i: :: :::::::::::::::::::::::

~)iii!:iig::ii{i!

!i .......... !i ......... i~ii~ii!!ii!~iiiiii:iii:i!!i~ili~ii~i~::~iii:~::_ :

i!. .~ .

::

:'iii~}}i~ti::: ............. ii'::ill

"''"'"'""~:

::::::::::::::::::::: :

Fig. 2. Geology of the South Wales coalfield and landslide distribution. Key: 1 Pantglas, 2 Troedrhiwfuwch, 3 Bournville, 4 New Tredegar, 5 Pentre, 6 Cwmtillery, 7 Victoria, 8 Ferndale, 9 Cumtillery Reservoir, 10 Darren ddu, 11 Cwm, 12 Troedrhiwgwair, 13 Blaina, 14 Duffryn, 15 Abercynon, 16 Morriston, 17 Pant-teg, 18 East Pentwyn, 19 Godre'r-graig, 20 Mynydd Corrwg Fechan, 21 Aberfan, 22 Mynydd Henllys, 23 Taren.

S=fe referred to

=' Landshdes

.

J:ii iiiii i!iii!ii|

Upper CoQ, He~lsures

~!!!!i!i!

~s

LEGEND

.

. . . . . :::::::;::::::Z:~============================ i ::;;:!:~ ii

.

.

i

.

t.

.

~

t,o

C~

24

AlMasmoum et al.

cyclothems are predominantly argillaceous in the Lower and Middle Coal Measures, which outcrop around the edge of the coalfield. In contrast, the Upper Coal Measures, or Pennant Series, are characterized by thick developments of sandstone which form the high central plateau of the coalfield. The drainage pattern, trending north-west to south-east, has been superimposed on the Coal Measures as a result of erosion of a cover of Mesozoic and Tertiary sediments, on which it originally developed. The valleys have been further incised by glaciers which extended into the coalfield on at least two occasions, finally retreating some 14 000 years ago. Presently, the slopes rise some 300 m from the misfit rivers to the plateau tops. Geological interest in South Wales grew as a direct function of the economic importance of ironstone and coal. Important areas of the coalfield were mapped by staff of the Geological Survey, H.T. De la Beche, W.E. Logan and D.H. Williams, and published as one inch to one mile maps (1 : 63 360 scale) in or before 1845. The first six inch to one mile maps (1 : 10 560 scale) were published around 1900. Important revisions to memoirs and maps, made in the early 1900s, were greatly facilitated by the readiness with which mining engineers and colliery officials supplied details. A second resurvey was started in 1945 and 36 1 : 10 560, 'county' maps were published between 1957 and 1960. When the post-war resurvey of the coalfield was completed in 1980, approximately 100 maps had been published. As urban renewal and improved infrastructure policies have been formulated, detailed geological information has been required to guard against geological and geotechnical hazards. Consequently the more recently published maps show landslides, and made ground and mine entries more exhaustively than earlier ones.

Formative years

Mapping in the coalfield, first by Logan and later by the Geological Survey, in the first part of the 19th century was almost entirely concerned with the solid geology. The only superficial deposit shown on the first edition maps was alluvium and no landslides were recorded. A letter from D.H. Williams to De la Beche, dated December 1840, describing the geology of Twmbarlwm, makes no reference to the ancient landslide in which Coal Measures are displaced over a 3 km section of Mynydd Henllys (Site 22 in Fig. 2). The first volume of the Memoirs of the Geological Survey of Great Britain, published in 1846 and including chapters on the geology and the denudation of South Wales, makes no reference to landslides. The six inch to one mile geological maps, published around 1900, were the first to include the annotation 'slipped ground' but the areas involved were not delineated. Few slip areas were shown. Taren (Site 23 in Fig. 2) in the Taft valley was marked but Mynydd Henllys was again overlooked. Clearly, sequential geological maps do not provide a basis for dating landslides. The earliest references to landsliding occur in newspaper reports of the 1860s and are concerned with their socio-economic significance. The Cwmbergwm mine was damaged by the Pantglas landslide in December 1860 (Free Press, 7 December 1860), trapping miners below ground. Seven months later, in the same area, occupants of a row of cottages narrowly escaped as rock falls and debris slides overwhelmed their properties (Free Press, 7 July 1861). A similar account is to be found of occupants evacuating their homes in the Rhymney Valley moments before it became engulfed in a 'mudslide' (The Cambrian, 1862).

Landslide research in South Wales

25

The first academic opinion to be published was that of Professor W.S. Boulton (Western Mail, 21 February, 1906). This article is important because it identifies the general public's perceptions of the causes of landslides; these include underground mine workings and a ubiquitous 'bed of quicksand'. Although Boulton set out to dismiss both of these factors, readers were presented with no convincing alternative explanations. The earliest reference in the scientific literature also appeared in 1906 (Galloway, 1906); this paper highlighted the association between landslides in the north-east part of the coalfield and the occurrence of 'coloured', Rhondda Beds (Red Beds). Public awareness of the potential risk associated with landslides became substantially increased as a result of events in the Rhymney valley. During the period 1901 to 1906 the New Tredegar colliery and the village of Troedrhiwfuwch, situated on the opposite sides of the valley, were subjected to a succession of damaging movements: at New Tredegar in August 1901, March 1905 and February 1906 and at Troedrhiwfuwch in March 1903 and February 1906. The events of February 1906 precipitated the articles by Boulton and by Galloway and reports claim that, as in the case of another slope stability disaster 60 years later, 'sightseers flocked to the area in their thousands' (Western Mail, 5 March 1906). Damage sustained to the New Tredegar colliery in 1905 appears to have made coal companies take a more cautious approach to ground movements on slopes. The awareness is evidenced by colliery drawings which defined the positions of 'new breaks' and 'moving zones' on slopes, and by mine plans, usually of adit workings, showing the position of landslips. There are numerous examples of remedial works in slopes adjacent to collieries, railways and roads; these usually took the form of drainage but occasionally walls and even rows of piles were used to retain or retard moving debris; it was also a common practice to monitor movements in order to judge the efficacy of remedial works. In addition to these examples of good practice there is evidence of malpractice, the most common of which was the excavation of material from the toe of landslides, usually for the purposes of constructing or maintaining roads and railways. Most landslide work would have been the responsibility of either mining engineers, acting for coal companies, or of local authority engineers; both groups must have gained some appreciation of causal factors and of the mechanics of landsliding although experiences were not reported in the literature until 1927.

Research 1916--1966

George Knox In 1916 Mr George Knox, a prominent mining engineer, acted as the principal scientific witness for the colliery company involved in the court case which followed the Pentre landslide. After two appeals, the case was settled in favour of the colliery company. The company had placed spoil in a quarry, partly overlapping a bank of quarry spoil, on the mountainside above the Rhondda village of Pentre. About two years later, after a further tier of tipping had been completed, a failure occurred in which colliery and quarry spoil slid down the hillside displacing colluvium into the backs of a row of cottages which had to be demolished. In 1927 Knox presented a 72-page paper titled, 'Landslides in the South Wales Valleys', to the South Wales Institute of Engineers. The full discussion and the demand for a second meeting reflects the considerable awareness of, and interest in, landslide problems which

26

AIMasmoum et al.

existed amongst engineers in South Wales at that time. Some information concerning the scale of the problem is provided in the documented discussion; Mr R.M. Evans, who appears to have represented Monmouthshire County Council, referred to 'the immense cost involved in repairing the damage done by landslides to railways, roads, water mains and sewers', and Mr W. O'Connor, another local authority engineer noted that 'it was a very timely paper as the problems in reference [sic] to local landslides were certainly on the increase'. The paper reveals an undercurrent of concern over the association between landslides and mining operations and one of the paper's principal objectives appears to have been to allay this concern; to this end Knox appears to have been successful. Mr R. Richards, a coal company representative, gives a noteworthy indication of the tenor of Knox's lecture; 'the facts put forward by Professor Knox must prove of exceptional value and guidance, and should largely, if not totally, dispel from the lay and professional mind the prevalent impression that the weight of colliery tips and movements due to subsidence were the principal contributory factors'. However, his arguments lacked objectivity. Knox's statement that, 'if a colliery tip.., rest on soil or rock liable to slide, it must slide with them, but it is in no way responsible for causing the slide', illustrates his inadequate appreciation of applied mechanics; interestingly this statement was challenged in the very first question of the discussion. A central theme to his paper was 'the fact that subsidence resulting from colliery workings must have a retarding effect on landslides, through drainage water into the mine workings'; by excluding the possibility that tilts and zones of extension may have detrimental effects on slope stability such remarks further illustrate Knox's dogmatism with respect to the consequences of mining. As proof of his contention, Knox quotes the example of New Tredegar where a shallow coal seam was worked with the specific intention of promoting drainage of the slope into the workings: 'the method adopted ... effectively stopped a large landslide moving towards the colliery'. For the record, these drainage measures were undertaken in 1912, mining of other seams continued until 1927, and in April 1930 New Tredegar colliery was destroyed by another major landslide. With respect to the consequences of mining, Knox himself could hardly be described as impartial. He was, at that time, Principal of the School of Mines in Pontypridd (later to become the Polytechnic of Wales, Treforest) and appears from his paper to have been advisor to several coal companies. Post-Knox hiatus

For the 40 years or so which followed Knox's contribution there was a hiatus in landslide research in South Wales. This was not due to an absence of landslides. In fact, some of the area's most serious landslides and flowslides of colliery waste were either initiated or were reactivated during this period (Table 1). These include the well-known flowside at Abercynon in 1939 which, because of its impact, prompted an investigation which was unusually extensive for that time. Despite these incidents, the area seems otherwise to have been of little interest either to academic researchers or to local and central government. Part of the explanation is quite clear. By 1927 Britain had entered the depression years, coal production was declining and any interest in landslides by coal companies would have declined as a function of their prosperity. On the part of local authorities there appears to have been little reaction apart from resignation; the belief, lingering from Knox, that landslides were, in effect, acts of

1984-1986

1984-1987

1987-1989

PECD/7/1/149-63/84

PECD/7/1/168

SWHP

SWHP/DMENU

SWHP

SWHP

IGS(EGU)

IGS(EGU)

IGS(EGU)

Investigation of causes of instability in the Blaina area, Ebbw Fach valley Survey of landslides in the Rhondda valleys and NE part of coalfield Regional survey of landslides in entire coalfield Investigation of controlling parameters at two, typical, shallow landslides; landslip susceptibility mapping; guidelines for site investigation in landslip areas Investigation of reasons for stability in an area between the active East Pentwyn and Bournville landslips, Blaina Development of 'landshp potential mapping' for use by planners and non-technical users Review of research into landsliding in UK, including distribution, causes, risk, remedial measures, legislative and administrative provisions Investigation of the tmpact of shallow, slow debris slides on buildings Research into the effects of mining on hillside stability and any links with 'first-time' deep-seated landslides Landslip potential mapping; monitored trial period of use

Research

Halcrow (1993a,b,c)

Halcrow (1989)

1991 University of Glamorgan (1993)

GSL (1987) RPT (1987)

Halcrow (1986)

Halcrow (1983)

Conway et aL (1983)

Conway et al. (1980)

Northmore et al. (1978)

Gostelow (1977)

Published

Extension of PECD/7/1 149-63/84

Includes Wales as one of six regional maps

Concurrent with ground investigations of the landslides for Gwent County Council

Extension of DGR482/45

Remarks

IGS (EGU), Institute of Geological Sciences (Engineering Geology Unit); SWHP, Sir William Halcrow & Partners; GSL, Geomorphological Services Ltd; RPT, Rendel Palmer & Tritton; BRE, Building Research Establishment; PW, Polytechnic of Wales, Treforest (now University of Glamorgan); DMENU, Department of Mining Engineering, Nottingham Umversity.

1988-1991

1983 & 1984-1985

PECD/7/1/093-102 PECD/7/1/178

PECD/7/1/285

1981-1983

F3/CR05/A/32/21

1987-1988

BRE/PW

1979-1980

482/87

PECD/7/1/256

GSL/RPT

1977-1979

DGR482/45

IGS(EGU)

1976

DGR482/17

Contractor

Contract period

Contract number

Table 1. DoE funded landslide research in South Wales

b~ "---I

28

AIMasmoum et al.

God must have been an important influence particularly since this view had been reinforced by the legal precedent of the Pentre landslide. And, of course, the Second World War must have done much to detract attention from incidents which now would have resulted in major inquiries. It is more difficult to understand why the hiatus in research continued after the War; in the 1940s and 1950s soil mechanics was developing into a major research discipline, and the Welsh valleys could have provided a wide variety of landslides for scientific investigation. Skempton (1946) referred to a New Tredegar landslide using one of Knox's illustrations, but made no mention of the much more damaging 1930 event, and he again mentioned South Wales landslides in a 1954 paper on the Jackfield landslide (Henkel and Skempton, 1954). Professor Skempton also visited East Pentwyn in 1954 to advise the local authority on the risk to property. Apart from these contributions, the scientific study of landslides in South Wales all but ceased.

Research 1966-1975

Aberfan

In 1970 there were over 500 spoil tips in South Wales. Situated on valley sides and plateau tops, their conical shapes punctuated the skylines of most valleys. At 9.15 a.m. on 21 October 1966 a flowside was triggered in the colliery spoil of No. 7 tip above the village of Aberfan. About one-third of the material in the flowside of 100000m 3 travelling at 16-32 km/h reached the village. The colliery waste engulfed the local junior school and 18 houses, killing 116 children and 28 adults. The Tribunal appointed to inquire into the disaster arranged for a series of scientific investigations to be performed under the general direction of Professor A.W. Bishop (Bishop et al., 1969). Aberfan provided an important focus for scientific research; advances made in the previous 20 years in the appreciation of shear strength (Skempton, 1964), pore pressures (Skempton, 1954), laboratory and in-situ testing (Bishop and Henkel, 1962), stability analyses (Janbu, 1954, Bishop, 1955), and aerial photo interpretation were applied to South Wales for the first time: the significance of mining-induced strains on the secondary permeability of sandstone aquifers became recognized (Wardell and Piggot, 1969). Collectively, these aspects formed a 'blueprint' for future investigations within the coalfield. On a national scale, the National Coal Board also initiated several other research projects into, inter alia, the geotechnical properties of spoil and lagoon deposits (MacKechnie-Thomson and Rodin, 1972; Bishop, 1973); the effects of subsidence and strain induced by undermining of spoil heaps (Forrester and Whittaker, 1974a,b) and liquefaction potential (Taylor et al., 1978). The importance of disasters lies in the lessons which can be learnt and the measures which can be taken to prevent their recurrence. Aberfan was no exception and far-reaching changes in the legal and managerial framework for colliery waste disposal resulted. Existing legislation was revised to embrace colliery and mine tips (Mine and Quarries (Tips) Act 1969) and general requirements for all existing and proposed tip sites associated with active mines and quarries were subsequently set out in the Mines and Quarries (Tips) Regulations 1971. The main thrust of the legislation was to ensure that spoil heaps were designed, constructed and maintained as any other engineering structure.

Landslide research in South Wales

29

Considerable importance was to be placed on the identification of geological and hydrogeological factors which could influence their security. Immediately after the disaster, but preceding legislation, a priority list was drawn up by the National Coal Board in the South Wales Area of tips which, by virtue of their potentially threatening locations, required immediate investigation. A Tip Control Unit was set up and preliminary geological reports were prepared with the assistance of senior field staff of the Institute of Geological Sciences. Soon afterwards, independent consulting civil engineers were appointed to assist the Board in South Wales in the first of several investigation programmes which eventually included over 250 tips and tip complexes. Of the sites investigated, about a quarter were found to have suffered instability either because of inherently weak foundation materials or because previous methods of tipping had led to unstable conditions. A considerable fund of data was assembled on the geotechnical characteristics of spoil and superficial deposits. By the time the project had been completed, some s million (present prices) had been expended on investigations and remedial measures. It can therefore be seen that Abeffan was a powerful stimulus to geotechnical investigation in the coalfield. It reawakened awareness of potential hazards of uncontrolled spoil disposal and created a particular sensitivity in the general public to problems of slope stability. It was a consequence of the spoil heap investigations in South Wales, when the stability of the underlying and adjacent terrains had to be considered, that a new generation of government agencies became aware of a 'natural' landslide problem.

The beginning of modern scientific research In the years immediately following Aberfan the general public's perception of the hazards of natural landslides is reflected by the newspaper accounts of 'threatening moving mountains' (South Wales Echo, 10 October 1967, 4 September 1969; Western Mail, 17 May 1973). Actual or potential instability resulted in the evacuation of three villages (Godre'r-graig, Fig. 3, Troedrhiwfuwch and Troedrhiwgwair) during the early 1970s. Geotechnical interest surfaced with a conference on 'Civil Engineering Problems in the South Wales Valleys' (Institution of Civil Engineers, 1970) in which Anderson and Trigg (1970) made reference to some of the larger landslides. Anderson had also initiated a programme of research into the glacial and post-glacial sediments within the coalfield valleys (Anderson, 1968, 1974) which remains a useful primary source of data for investigations today. It can be argued that scientific analysis of natural landslides in the coalfield by geomorphologists, as 'academic' research, pre-dated more positive activities prompted by practical necessity. Slope stability models based on soil mechanics principles were being investigated as early as 1965 in West Glamorgan (Rouse, 1969) and later in the Lliw Valley (Rouse and Farhan, 1976). Frequency distributions of randomly sampled slopes were interpreted using an infinite slope model to suggest that shallow mass movement processes had been important in their development. The early 1970s saw a wider recognition that many fossil shear surfaces found within colluvium on the valley sides were created as a result of these processes within the periglacial climate which existed around the margins of the retreating Devensian valley glaciers. Their affinities with solifluction processes in present day periglacial environments were described by Chandler (1972) who drew on an example exposed at East Pentwyn. The existence of shallow fossil shear surfaces has occasionally proved troublesome in the redevelopment of the valleys, particularly during

30

A l M a s r n o u m et al.

Fig. 3. Godre'r-graig and Pant-teg landslides (by kind permission of the Welsh Industrial and Maritime Museum) road construction when surface manifestations of landslide features have not always been obvious. Reinterpretation of some of the slides described by Knox suggests that tipping on pre-existing shear surfaces within colluvium was their main cause. Morphological and geomorphological mapping has evolved into an important technique in ground investigations for a wide range of civil engineering operations; one of the first applications of this technique was in the selection of a route for the A470 Taft Vale Trunk Road (Brunsden et al., 1975). Similar techniques are now routinely used to define the processes within, and development of, individual landslides.

Research 1976 to the present Since the mid-1970s, there have been great advances in the scientific study of landslides in South Wales through research sponsored by the Department of the Environment (DOE). The reasons for this research are firmly embedded in the increased awareness of geotechnical problems which followed Aberfan. Social and political pressures to tackle certain potentially threatening 'natural landslides' such as those at Blaina in the north-east

Landslide research in South Wales

31

of the coalfield, arose which created financial, administrative and legal problems for the local authorities concerned. At one time, it was suggested that a 'Geotechnical Centre for Wales' be established to allow the co-ordination of field and laboratory studies of landslides and associated geotechnical problems. Nothing came of this initiative because it was argued that existing research centres such as the Transport and Road Research Laboratory, Building Research Establishment and Institute of Geological Sciences (IGS) were capable of carrying out the work. DoE research on the landslides in South Wales began in the mid-1970s with the transfer of funds from Natural Environment Research Council/IGS to customer departments and the establishment of the customer-contractor principle. For the first few years of the new arrangements, Welsh landslip research formed part of a larger programme of engineering geology research carried out and largely initiated by the Engineering Geology Unit of the IGS and managed by the DoE. In the early 1980s, management of individual projects passed to the Welsh Office, although the overall organization of the programme was a joint DoE/Welsh Office exercise. Table 1 lists the projects which have been carried out and illustrates the many advances that have been made. Not surprisingly, the first project was site specific and was aimed at establishing the reasons for widespread instability in the area of Blaina and of the East Pentwyn and Bournville landslides in particular (Fig. 4). The project was performed by Dr T.P. Gostelow of the IGS, who brought to the coalfield considerable experience gained through

Fig. 4. Bournville landslide (by kind permission of the Welsh Office)

32

A1Masmoum et al.

working on coastal landslides in south-east England with Professor J.N. Hutchinson. Although some aspects of the landslide processes and hydrogeology were later found to be inaccurate, Gostelow (1977) reiterated the factors originally recognized by Knox which made the area susceptible to movement. Unlike Knox, however, Gostelow concluded that mining had promoted movement by altering groundwater flow paths (see Daughton et al., 1977) inducing ground tilts and reactivating faults. Subsequent projects followed a general policy which has evolved into three phases: (a) regional surveys; (b) investigation of the causes and mechanisms of landslides; (c) assessment of landslip potential.

Regional surveys

In order to establish the distribution of landslides in the north-eastern and central parts of the coalfield, surveys were carried out in 1977-1979 (Northmore et al., 1978). Between 1979 and 1980 the survey was enlarged to encompass the entire coalfield. The resulting South Wales Coalfield Landslip Survey (Conway et aL, 1980) is a two-volume report accompanied by 1 : 50 000 scale maps showing landslides in relation to topography, solid geology, superficial deposits and slope angles. In the main volume of text, the geological, hydrogeological and geotechnical controls on slope instability were described, together with a classification of landslide types in terms of depth (shallow/deep) and style of movement (rotation/translation/flow). The classification owed much to that used by Varnes (1958) in which the term 'flow' was used to describe that type of shallow translational movement of soil or rock debris now recognized as a slide, i.e. with a discrete sheared surface at its boundary (Hutchinson, 1970). The second volume is a catalogue describing the main features, classification, activity and area of each of the 579 landslides identified. Typical extracts from the catalogue are reproduced in Table 2. The publication of the South Wales Coalfield Landslip Survey was a major advance for those local government officers, consulting engineers and academic researchers interested in landslides. For the first time, the extent and main characteristics of landslides in South Wales had been documented and provided the basis for later research targeted at those aspects of landslide management or more relevance to local authorities and developers. However, the preparation of chronological inventories of landslide events and damage sustained was not included as requirements in the regional-survey contracts. Trends in the scale of the landslide problem were not, therefore, defined.

Investigation of causes and mechanisms

Having established the main categories of landslide processes found within the coalfield, one branch of subsequent research concentrated on specific sites considered representative of some of the main landslide types. Conway et al. (1983) carried out detailed investigations to establish the causes and mechanisms of two shallow landslides of the type which represented 93% of all those recorded in the South Wales Coalfield Landslip Survey. They demonstrated, inter alia, that the stability of the two landslides was governed by perched water table conditions. A close correspondence between r values determined from laboratory ring shear tests and values determined by back-analyses indicated that strengths at or close to residual were being mobilized on the shear surfaces

E side of the Ebbw Fach valley above E Pentwyn Farm, W of Blaena.

W side of the Ebbw Fach valley below Darren Ddu.

W side of the Ebbw valley, above Victoria.

EB 67

EB 77

Localion

EB19

Slip number

SO 170 072

SO 200 060

SO 207 075

National Grid reference

Tf.

XT.

DRFx. Deep-seated rotational slip grading downslope into debris slides and flows. Slip in Llynfi, R h o n d d a and Brithdir Beds. A n accumulation of rockfall debris below a steep Hughes sandstone backscar above the Cefn Glas coal seam. Debris is now active as a translational rockslide with associated minor surface ' m u d ' flows at the toe. Active rockfall is still occurring. Shallow translational debris slides and flows, mainly in superficial material and Llynfi shale, below the outcrop of the No. 2 R h o n d d a coal seam. The slip also extends a little above the No. 2 seam into the Rhondda Beds.

Summary description

Table 2. Typical extracts from South Wales Coalfield Landslip Survey (Conway et al. 1980)

A

A

A

Activity

320 280

350 240

450 330

(m)

O D height top bottom

Dimensions

110

300

350

Downslope length (m)

Slip area (hectares)

r~

r~

W side of the Sirhowy valley above the village of Bedwellty Pits.

SR 11

SO 155 060

SO 136 050

National Grid reference

Fr.

DRtfx.

Deep-seated rotational slip in Rhondda Beds with minor translational debris slides and flows below. Some movement in 1905 followed by major activity from 1925 to 1930. Current activity is confined to rockfall from the backscar and minor movement at the toe. Shallow flows in superficial material and Rhondda shale with minor rotational failures. The slip has occurred below the outcrop of the Brithdir coal seam and is, in part covered by spoil from workings in this seam. Current minor flow activity occurs in the south-eastern part of the slip.

Summary description

A

A

Activity

335 274

350 240

OD height top bottom (m)

Dimensions

300

300

Downslope length (m)

16

25

Slip area (hectares)

Landslip types: m a j o r m o v e m e n t s - D R deep-seated rotation al, T translational, F flow, X fall; m i n o r m o v e m e n t s - t translational, r rotational, f flow, x fall. Valleys: EB Ebbw, R H Rhymney, SR Sirhowy. Activity: A active, D dormant.

New Tredegar on the E side of the Rhymney valley between New Tredegar and Abertysswg.

Location

RH 3

Slip number

Table 2. (Continued)

-~

Landslide research in South Wales

35

within these shallow landslides. The project also established guidelines for site investigations in other landslide areas. Investigations have been applied to deep-seated landslides and in particular the role of mining in the initiation of first-time slides (Franks, 1983; Franks et al., 1985). State-of-theart techniques of physical and finite-element modelling of critical phases of mining at certain sites have demonstrated the importance of mining strains in promoting movement (Jones et al., 1992). Assessment of lands@ potential It has been argued that the level of natural landslide hazard in South Wales is lower than assumed previously as most landslides can be ascribed to the deleterious effects of human interference at old landslide sites. It is not surprising that many damaging landslide events date from the period when the coalfield valleys were being intensively developed, mostly in ignorance of ground conditions. A methodology by which landslide hazard could be portrayed in a form suitable for use in planning and for development control was therefore identified as an important research area. Pilot studies for such a methodology were performed by Forster (1982) and by Conway et aL (1983). In both cases, hazard zonation was achieved by superimposing maps of solid geology with those of slope angle and modifying the result to take account of geomorphology. It was concluded that landslide susceptibility mapping in South Wales was possible and could offer benefits to local planners and to those involved in feasibility studies for construction projects. The pilot studies were considered sufficiently encouraging for the Welsh Office to initiate a research project for trial landslip potential mapping of 60 k m 2 of the Rhondda Valleys (Halcrow, 1986). The methodology involved in determining the degree of association between landslide locations and variations in the attributes of those factors considered to influence them. Numerical ratings were derived for each attribute and then combined to define 'landslip potential' as a number (Payne, 1986). The basic parcels of land used in the preparation of the 1 : 10 000 landslip potential maps were 50 • 50 m elements (Fig. 5). This allowed computerized methods of data storage, and manipulation, and map production. Mapping has since been extended to the entire administrative area of Rhondda Borough. Research also included the examination of ways in which the maps could be implemented in the planning process (Siddle et aL, 1987) and by the provision of 'Planning Guidelines' for a period of trial use of the maps.

Landslide management In South Wales, and in other areas of the UK with histories of slope movements, landslide management falls within the remit of many organizations, each of which requires different types of information to fulfil its role. However, before landslide management on a regional basis can be justifiably resourced it is necessary to quantify the problem. This should involve an assessment of landslide events throughout recent history to provide an inventory of their type, size and impact. The degree of hazard (probability of occurrence) and risk (degree of damage) can thus be evaluated with chronological perspective. Against this background of data, long-term resource allocation can be planned and priority areas identified.

C>

II II

II -

-

5~B

240

landslip

Br

LANDSLIP LANDSLIP

POTENTIAL AREA

POTENTIAL

~HONDDA

HAF

ASSESSMENT

Welsh Of(ics

18~d6lt~

PLANNING

O~par~n~t Of ~r~e Env~on~=nt

~r

e~r165

A P ~ wlt~ ~eme f ]an~5~1~ ~otentl~l

O~ Bcti~e ~nO

~40

l~i

33 - ~25

A.ear =f

~Pe~E

~P

[~~

~escP~pE~On

L~

Landslide research in South Wales

37

In the first stage of landslide research, per se, causal processes and failure mechanisms for each landslide type should be identified and some assessment made of whether the causal processes change with time. Secondly, hazard maps are required, portraying factual data on existing landslides and predictive data concerning landslide susceptibility (or potential) and future incidence. These maps need to be at scales suitable for application to development control and planning. Thirdly, risk maps should be prepared, by combining data on hazard and land use, in order to assess likely future impact, based on historical trends. These enable the cost-effectiveness of various management strategies to be evaluated and decisions made on their implementation. Risk maps of this kind have particular application to the management of active landslides which threaten property. Lastly efforts need to be made to communicate the results of the research to those who would benefit from it. This involves interfaces between the researcher and the various end users of the information. This 'ideal' pattern of activity has not been followed in South Wales: rather research has been based on the perception of a high probability of damaging landslide events rather than scientific inquiry.

Conclusions The development of research in any discipline is usually the result of practical need and academic interest. This paper has demonstrated that the problems created by landsliding in South Wales have long been recognized and that the measures required to stabilize or avoid them being discussed with some seriousness as long ago as 1927. However, research interest in landslides dwindled for some 40 years before being reawakened in the aftermath of the Aberfan disaster. It is evident that the concentration on South Wales for the major proportion of government-funded landslide research can be traced to the increased awareness of slope stability problems which followed Aberfan and to the consequent political pressures by some local authorities for assistance to tackle them. The perceived degree of hazard within the coalfield was made largely independently of any comparison with other areas of the UK. This has only recently been made possible by the compilation of a database for all recorded landslides in Great Britain (Geomorphological Services Ltd, 1987). It may be that, had this information been made available earlier, research may have targeted to other areas which have been shown to possess even greater densities of landsliding and where risks to development are more severe. Although the earliest research appears to have been directed at specific landslides, subsequent studies were broad-based and have resulted in a framework of information and techniques that local authorities and various developers can apply to new slope problems.

Acknowledgements For their helpful comments the authors thank Dr D. Brook of the Department of the Environment, Mr H.R. Payne of the Welsh Office, and Dr R.A. Bazley, British Geological Survey at Aberystwyth.

38

A l M a s m o u m et al.

References

Anderson, J.G.C. (1968) The concealed rock surface and overlying deposits of the Severn valley and estuary from Upton to Neath, Proceedings of the South Wales Institute of Engineers, 83, 27-47. Anderson, J.G.C. (1974) The buried channels, rock floor and rock basins, and overlying deposits, of the South Wales valleys from Wye to Neath, Proceedings of the South Wales Institute of Engineers, 88, 1-17. Anderson, J.G.C. and Trigg, C.F. (1970) Geotechnical factors in the redevelopment of South Wales valleys, Proceedings of the Conference on Civil Engineering Problems of the South Wales valleys, Institution of Civil Engineers, Cardiff, pp. 13-22. Bishop, A.W. (1955) The use of the slip circle in stability analysis of earth slopes, Geotechnique, 5, 7-17. Bishop, A.W. and Henkel, D.J. (1962) The Triaxial Test, Edward Arnold, London. Bishop, A.W., Hutchinson, J.N., Penman, A.D.M. and Evans, H.E. (1969) Geotechnical Investigation into the Causes and Circumstances of the Disaster of 21st October 1966, A Selection of Technical Reports Submitted to the Aberfan Tribunal, HMSO, London, pp. 1-82. Bishop, A.W. (1973) The stability of tips and spoil heaps, Quarterly Journal of Engineering Geology, 6, 335-77. Brunsden, D., Doornkamp, J.C., Fookes, P.G., Jones, D.K.C. and Kelly, J.M.H. (1975) Geomorphological mapping techniques in highway engineering, Journal of the Institute of Highway Engineers, 21, 35-41. Chandler, R.J. (1972) Periglacial mudslides in Vestpitsbergen and their bearing on the origin of fossil 'solifluction' shears in low angled clay slopes, Quarterly Journal of Engineering Geology, 5, 223-41. Conway, B.W., Forster, A., Northmore, K.J. and Barclay, W.J. (1980) South Wales Coalfield Landslips Survey, IGS Special Surveys Division, Engineering Geology Unit, Report, 80/4. Conway, B.W., Forster, A. and Northmore, K.J. (1983) A Study of Two Landslipped Areas in the South Wales Coalfield, IGS Engineering Geology Unit, Report, 83/6. Daughton, G., Noake, J.S. and Siddle, H.J. (1977) Some hydrogeological aspects of hillsides in South Wales, Proceedings of the Conference on Rock Engineering, Newcastle-Upon-Tyne, pp. 423-39. Forrester, D.J. and Whittaker, B.N. (1974a) Effects of mining subsidence on colliery spoil heaps: Part 1, Mining subsidence and geotechnical aspects of spoil heaps, International Journal of Rock Mechanics & Mining Science, 13, 113-20. Forrester, D.J. and Whittaker, B.N. (1974b) Effects of mining subsidence on colliery spoil heaps: Part 2, Deformation behaviour of spoil heaps during undermining. International Journal of Rock Mechanics & Mining Science, 13, 121-33. Forster, A. (1982) South Wales landslip hazard assessment. A pilot study, IGS Engineering Geology Unit, 82/6. Franks, C.A.M. (1983) An analysis of the effects of hillside topography on mining subsidence in the South Wales Coalfield. PhD thesis, University of Wales. Franks, C.A.M., Bentley, S.P. and Geddes, J.D. (1985) Mining histories and their effect on slope stability, Proceedings of the Symposium on Landslides in the South Wales Coalfield, Polytechnic of Wales, pp. 207-24. Galloway, W. (1906) The landslide in the Rhymney valley, Nature, 73, 425-6. Geomorphological Services Ltd (1987) National Review of Landsliding in Great Britain. Series A, IV, Wales, Report to the Department of Environment, London. Gostelow, T.P. (1977) The development of complex landslides in the Upper Measures of Blaina, South Wales, Proceedings of the Symposium on the Geotechnics of Structurally Complex Formations, Capri, pp. 225-68. Halcrow, Sir William, & Partners (1985) East Pentwyn and Bournville Landslips Research Project, Report for Department of the Environment and Welsh Office. Halcrow, Sir William, & Partners (1986) Assessment of Landslip Potential: South Wales (Rhondda

Landslide research in South Wales

39

Valleys Landslip PotentialAssessment), Report for Department of the Environment and Welsh Office. Halcrow, Sir William, & Partners (1989) Landslides and Undermining Research Project, Report for Department of the Environment and Welsh Office. Halcrow, Sir William, & Partners, (1993a) Rhondda Landslip Potential."Planning Guidelines, Report for Department of the Environment and Welsh Office. Halcrow, Sir William, & Partners (1993b) Rhondda Landslip Potential."Summary Report, Report for Department of the Environment and Welsh Office. Halcrow. Sir William, & Partners (1993c) Rhondda Landslip Potential: Final Report, Report for Department of the Environment and Welsh Office. Henkel, D.J. and Skempton, A.W. (1954) A landslide at Jackfield, Shropshire, in an overconsolidated clay. Proceedings of the European Conference on Stability of Earth Slopes, Stockholm. Vol. 1, 90-101. Hutchinson, J.N. (1970) A coastal mudflow on the London clay cliffs at Beltinge, North Kent, Geotechnique, 24, 412-38. Institution of Civil Engineers (1970) Civil Engineering Problems of the South Wales Valleys, Institution of Civil Engineers, London. Janbu, N. (1954) Application of composite slip circles for stability analyses, Proceedings of the European Conference on the Stability of Earth Slopes, Stockholm, Vol. 4, 43-49. Jones, D.B., Reddish, D.J., Siddle, H.J. and Whittaker, B.N. (1992) Landslides and undermining: slope stability interaction with mining. Proceedings of the 7th International Society of Rock Mechanics Congress, Aachen, 893-8. Knox, G. (1927) Landslides in South Wales valleys, Proceedings of the South Wales Institute of Engineers, 43, 161-247, 257-90. McKechnic-Thompson, G. and Rodin, S. (1972) Colliery spoil tips after Aberfan, Proceedings of the Institution of Civil Engineers, 7522, 1-59. Northmore, K.J., Duncan, S.V. and Gostelow, T.P. (1978) Landslip Survey of the South Wales Coalfield Valleys - Rhondda and North-east Survey Areas Interim Report, Institute of Geological Sciences, Internal Report, 78/17. Payne, H.R. (ed.) (1986) Assessment of Landslip Potential: South Wales. Summary of the Research and Description of the Mapping Technique Developed, Welsh Office, Cardiff. Rendall, Palmer and Tritton (1987) A review of investigation techniques and remedial measures, research and practice, in, National Review of Landsliding in Great Britain, Series C, Report for the Department of the Environment. Rouse, W.C. (1969) An investigation of the slope stability and frequency distribution of slopes in selected areas of West Glamorgan. PhD thesis, University of Wales. Rouse, W.C. and Farham, Y.I. (1976) Threshold slopes in South Wales, Quarterly Journal of Engineering Geology, 9, 327-38. Siddle, H.J., Payne, H.R. and Flynn, M.J. (1987) Planning and Development Control in an Area Susceptible to Landslides, Planning and Engineering Geology, Geological Society Engineering Geology Special Publication No. 4, 247-53. Skempton, A.W. (1946) Earth pressure and the stability of slopes, in The Principles and Application of Soil Mechanics, Institution of Civil Engineers, pp. 31-61. Skempton, A.W. (1954) The pore pressure coefficients A and B, Geotechnique, 4, 143-7. Skempton, A.W. (1964) Long-term stability of clay slopes, Geotechnique, 14, 77-101. Taylor, R.K., Kennedy, G.W. and Macmillan, G.L. (1978) Susceptibility of coarse grained coal-mine discard to liquefaction, Proceedings of the 3rd International Congress, b~ternational Association of Engineering Geologists, Madrid, pp. 91-100. University of Glamorgan Commercial Services (1993) A Study of a Slow-moving Landslide and the Effect on a Local Community, its Buildings and Infrastructure, Report to the Department of Environment and Building Research Establishment.

40

AlMasmoum et al.

Varnes, D.J. (1958) Landslide types and processes, in Landslides and Engineering Practice, Eckel, E.B. (ed.) Highway Research Board Special Report 24, 20-47. Wardell, K. and Piggott, R.J. (1969) Report on Mining Subsidence. A Selection of Technical Reports Submitted to the Aberfan Tribunal, HMSO, London, pp. 187-204. Woodland, A.W., Evans, W.B. and Stephens, J.V. (1957) Classification of the Coal Measures of South Wales with special reference to the Upper Coal Measures, Bulletin of the Geological Survey of Great Britain, 13, 6-13.