BRITISH GEOLOGICAL SURVEY

Natural Environment Research Council

Mineral Reconnaissance Programme

Report No. 92 A mineral reconnaissance survey of the Llandrindod Wells/Builth Wells Ordovician inlier, Powys

Geochemistry T. R. Marshall MSc R. C. Leake BSc, PhD

Geology T. R. Marshall MSc

Geophysics K. E. Rollin BSc

0 Crown copyright 1987

A report prepared for the Department of Trade and Industry Keyworth, Nottinghamshire 1987

CONTENTS FIGURES

Summary 1 la Geography of Builth Wells Inlier 2 lb Structural location of Builth Wells inlier 2 Introduction 1 2 Distribution of boron in sieved stream sediment Geography 1 samples 6 3 Distribution of calcium in panned concentrate Geology 1 samples 7 Stratigraphy and lithology 1 4 Distribution of titanium in panned concentrate Structure 3 samples 8 Superfkial deposits 4 5 Distribution of vanadium in sieved stream Previous mining activity 4 sediment samples 9 MRP Reconnaissance Exploration 4 6 Distribution of manganese in sieved stream Geochemical sampling 4 sediment samples 10 Chemical analysis 4 7 Distribution of iron in sieved stream sediment Treatment of data 4 samples 11 Drainage reconnaissance 5 8 Distribution of iron in panned concentrate Boron (sieved sediment) 5 samples 12 Calcium (panned concentrate) 5 9 Distribution of cobalt in sieved stream sediment Titanium (panned concentrate) 5 samples 13 Vanadium (sieved sediment) 5 10 Distribution of copper in sieved stream sediment Manganese (sieved sediment) 10 samples 14 Iron (sieved sediment) 10 11 Distribution of copper in panned concentrate Iron (panned concentrate) 10 samples 15 Cobalt (sieved sediment) 12 12 Distribution of zinc in sieved stream sediment Copper (sieved sediment) 12 samples 16 Copper (panned concentrate) 12 13 Distribution of zinc in panned concentrate Zinc (sieved sediment) 15 samples 17 Zinc (panned concentrate) 15 14 Distribution of strontium in panned concentrate Strontium (panned concentrate) 17 samples 18 Zirconium (sieved sediment) 19 15 Distribution of zirconium in sieved stream Tin (panned concentrate) 19 sediment samples 19 Barium (sie\.ed sediment) 19 16 Distribution of tin in panned concentrate Barium (panned concentrate) 19 samples 20 Cerium (panned concentrate) 22 17 Distribution of barium in sieved stream sediment Lead (sieved sediment) 22 samples 2 1 Lead (panned concentrate) 24 18 Distribution of barium in panned concentrate Synthesis of reconaissance drainage data 24 samples 22 19 Distribution of cerium in panned concentrate Follow-up drainage sampling 26 samples 23 Reconnaissance soil sampling 27 20 Distribution of lead in sieved stream sediment samples 24 Detailed soil sampling 35 21 Distribution of lead in panned concentrate Detailed geology 36 samples 25 Structure 37 22 Synthesis of distribution of base metal anomalies in drainage samples 26 Geophysical surveys 37 23 Location of follow-up grab sampling drainage Drilling 37 sur\reys 27 24 Distribution of Pb and Zn anomalies in follow-up Mineralisation 42 drainage surveys 28 Conclusion 43 25 Plot of Zn, Pb, Cu, As, Ni, V, Cr, Mn, Fe and Loss on ignition (LOI) 1evels in reconaissance soil Acknowledgements 44 line 1 29 References 44 26 Distribution of Cu, As, Zn and Pb soil anomalies (Cu, As point, Zn, Pb contoured) around the area of old mining activity between Llandrindod Hall and Llwynceubren Farm 30 27 Distribution of Cu anomalies in reconnaissance soil lines 31 28 Distribution of Zn anomalies in reconnaissance soil lines 32 29 Distribution of Pb anomalies in reconnaissance soil lines 33 30 Distribution of Fe, Mn, Pb and Zn anomalies (contoured) over northern part of main tuff group 34 31 Geology of northern part of main tuff group 35

. . . 111 32 Contoured plots of chargeability and apparent 37 Section to show interpretation of relationship resistivity together with Pb and Zn in soil around between surface soil anomalies and Pen Rhiw Pen Rhiw Frank. Borehole locations also Frank borehole 3 43 shown 36 33 Photogeological map of area around Pen Rhiw Frank with borehole locations 38 TABLES 34 Compilation of results of geophysical surveys over northern part of outcrop of main tuff group 39 1 Stratigraphical subdivisions in Llandrindod Wells 35 Triangular CaO-MgO-K,O plot of borehole area 3 samples in sequence down each hole 40 2 Analytical techniques utilised in exploration 5 36 Graphic log and plot of MgO, SiO,, K,O, CaO, 3 Summary statistical characteristics of each element Fe,O,, S, Pb and Zn contents of Pen Rhiw Frank in drainage samples 5 borhole 3 41 4 Range of elements in boreholes 42 SUMMARY any stratabound mineralisation at depth below the zone of surface Lveathcring. A further consequence of the flat dip A reconnaissance drainage geochemical survey was car- of strata is that by accidents of topography extensive over- ried out at a density of about 2 samples per km2 of the burden anomalies could be produced from relatively Llanclrindod Wells/ Builth Wells Ordovician inlier in minor statabound base-metal enrichments on the hill Powys. The rocks of the area comprise a sequence of tops. mudstones and shales ranging in age from the Llan\.irn to the Caradoc containing volcanic horizons and a lrolcanic complex of acid/intermediate composition overlain in the INTRODUCTION south by a further volcanic sequence of more basic com- position. An area of about 65 km2 was sampled and from Drainage and soil surveys were sampled in 1978 and 1979 each site a wet screened minus BSI 100 mesh (< 150 o\‘cr the outcrop of volcanic and sedimentary rocks micrometres) fraction and panned heavy mineral concen- known as the Llandrindod Wells (or Builth Wells) Or- tratc samples were obtained. Sie\rcd sediment and con- do\.ician inlier. There is no evidence of extensive base- centratc samples were analysed for B, V, Cr, Mn, Fe, metal vein mineralisation such as found in the Shelve Or- Co, Ni, Cu, Zn, As, Zr, MO, Ag, Sn, Ba and Pb and Ca, do\.ician Inlier to the NNE or throughout the Central Ti, Mn, Fe. Ni, Cu, Zn, Sr, Zr, MO, Sn, Sb,Ba, Ce, Pb Wales Mining Field. Nevertheless by analogy with and U respecti\.ely. Several Pb and Zn anomalies were mineralised \.olcanic belts throughout the world the area apparent many of which appeared to be associated with was judged worthy of examination for metallic the outcrop of the main volcanic unit of the area and 4 of mineralisation especially of the stratabound variety. The these were confirmed by further grab stream sediment area is included in the 1:50 000 topographical sheet 147 and bank sampling. Drainage sampling was sup- (Elan \‘alley and Builth Wells) and is covered by a special plemented by 27 reconnaissance soil tra\rerscs mostl) 1:25 000 gological map in the BGS series ‘Classical areas across the main outcrops of the \.olcanic- rocks which tencl- of British Geology’ (Earp, 1977). ed to form the most ele\rated ground. The soil samples Geochrmical drainage data are presented in map form were sie\.ed to give a minus 85 BSI mesh (< 180 (Figures 2 - 23) illustrating the distribution of 15 elements micrometres) sample for analysis for Cu, Pb, Zn, Ni and in minus 100 BSI mesh (150 microns) stream sediments either Mn and Fe or Ag and Co. Samples collected from and panned hea\.y mineral concentrates. An area of an art‘s of‘old lcaci workings in the west of the inlier failed 65 km was covered with a sample density of approximate- to pro\ridc c\icience of further vein mineralisation beyond ly two samples per square kilometre. Variation in the limits of the workings. Elswhere the biggest concen- topography and drainage resulted in a lower sampling tration of‘ Pb and Zn anomalies and the greatest density o\‘cr the \,olcanic rocks, which were of potential amplitude of anonlaly (2000 ppm Pb) wcrc located over economic interest, than was desirable. Accordingly a the northern outcrop of the the main unit of tuffaceous number of supplementary reconnaissance soil traverses volcanic rocks, particularly in the \.icinity of the farm Pen were sampled, particularly o\rer the volcanic rocks. These Rhiw Frank. art‘ presented either as line profiles or contoured plans for A clctailccl geological, geochemical and geophysical elements of interest. survey of the Pen Rhiw Frank area was carried out using photogeological interpretation, further soil sampling and a combination of IP/resistivity, VLF, SP and magnetic GEOGRAPHY surveys. Geological interpretation indicated that the area was divided into fault blocks and the geophysical work The surlpey area (Figure 1) forms a NNE-trending belt of re\realecl a zone‘ of high resistivity with roughly coincident moderately rugged hills some 12 km long rising above the low amplitude chargeability maxima and a VLF cross- surrounding lowlands. The area itself can be divided into o\.t‘r o\c‘r part of the zone. There was also a general hills underlain by lavas and tuffs and lower undulating association of this zone with lead in soil anomalies and the terrain underlain by shale sequences. The highest areas presence of significant amounts of weathered pyrite in the arc on the Carneddau (400 m) in the south and Gilwern limited outcrops. The lead in soil anomalies when con- Hill (380 m) in the centre of the area but the most rugged toured in the light of the photogeological interpretation area is around Llandegley Rocks at the NE margin of the can by sc‘cbn as broadly following the local strike of the area. l’he area is bounded to the north by the River Ithon rocks. Four holes were drilled to in\restiage the source ot which drains westward into the Ri\rer Wye and cuts across the> geochemical anomalies and to investiage any associa- the regional geological structure. To the east the area is tion bt~tn~ecn this and the obser\red geophysical drained by the River Edw and Camnant Brook while the anonlalies. Thy holes wcrc sited to test the down dip ex- Ri\.cr b’ve tlows close to the SW boundary. tension of‘ apparently stratabound surface soil anomalies. ‘l’hc ce’ntres of population are the former spa towns of l’uff~accous rocks of andcsitic composition with a horizon Built h m’t,lls and Llandrindod Wells while the main oc- of clacitic. tuffs and minor lrcsicular andcsitic 1aL.a wt‘rc in- cupation of the area is farming. Much of the land is tcrsccted in the holes. Mineralisation occurs in the for-m of co\,crcd \vit h pasture or rough grazing. secondary lc~ad minerals within a poorly rcco\rcred soft clay-rich section 5.8 m thick near the top of ant‘ of the GEOLOGY holes. 1~~1 lc\.els up to 0.52 %I o\‘t‘r 3.4 m were found in the zone which is located at the interface between a dacitic Stratigraphy and lithology tuff unit chcn~icnlly distinct from the rest of the \.olcanic scqucrlcXX and an andesitic la\,a probably of linlitcd c\;- ‘I’hc art‘;1 hits long attracted the attention of geologists tent .‘I’hc (h-ill holes suggest that thcl dip of the strata is alt tlou~gh disco\~erics ot riiirieralisat ion of potential around 5 dcgrccs rather than the 30 degrees d~cluc~d from clconomic intt‘rcst ha\zc not been reported. Studies by th<> t,xisting gc,ological map of the arca and so the holes Murctiison ,l,apworth, CYoods ,md Elles, predated a wcr(‘ not sitccl at locations which wc’rc’ likcsly to iI1tc‘rsc‘c.t writs 01‘ p;lpc’rs t>v , .Jones and Pugh (1941, 1946, 1948, (LIansantffraed? 2 fiq ‘12 ‘? r/ ‘? _. * KEY

Z=ZIC Roads A class

North + Small Villages r_ I I

Figure la Geography of Builth Wells Inlier

_ _ _ Post Silurian

Post Ordowcian

Figure lb Structural location of Builth Wells inlier etc.) Lvhich M’ere based on mapping at the scale of 25 in- elastic sediments. The pyroclastic rocks are usually ches to the mile. These maps were simplified to the unwelded pumicious lappilli tuffs with common lithic 1: 10 560 scale and compliled into a 1:25 000 map of‘ the fragments of lava and shale. Invididual units range from inlier by Earp (1977). I n addition there has been a more 5 cm to 30 m in thickness and may occur as graded units recent study of‘ the volcanics by Furnes (1978). l’here or homogeneous units. They have been interpreted by ha\.e also been regional accounts of’the geology and evolu- Fumes ( 1978) as subaqueous pyroclastic flows. Welded tion of the W’elsh Basin (Kokelaar and others, 1984, tuffs are subordinate to unwelded tuffs and occur as Woodcock 1984 and BeLTins and others, 1984) which in- 4-30 m thick heterogeneous units, dominated by pumice clude limited data and discussion about the Llanclrinclod and \itric fragments. Although these units show many Wells/Builth Wells Ordovician inlier and its \*olcanic se- features considered reliable indicators of a subaerial en- quencv. W’hilc use has been made of Furnes’s descriptions \.ironment of deposition, Francis and Howells (1973) anal interpretation of the rocks base maps ha\Te been recognized submarine ignimbrites in the Cape1 Curig prepared usins the geological units and boundaries of the Formation of North Wales and Sparkes and others (1980) 1:25 000 BGS map. A stratigraphical column f’or the area ha\re obser\red pyroclastic flows which have been emplac- is gi\ren in 7’able 1. Shales above Main tuff group are col- cd and welded offshore from Dominica. These pass up- lt‘cti\.c‘ly known as Upper Shales and shales below Main \varcls into basic la\ras which are well displayed at tuff‘ Group arc‘ known as Lower Shales. I,lanel~vcdd Quarries [SO 049 5221 and have been described by Nichols (1956). The erosional top shown by one basic flow taken in conjunction with the general Table 1 Stratigraphic subdivisions in Llandrindod absence of hyaloclastites or pillow formation strongly sug- Wells area gcsts subaerial eruption and emplacement of the basic flows. Acid la\ras are common throughout the main tuff group and arc shown on the BGS 1:25 000 geological map Tuff and agglomerate as ‘rnassi\~e keratophyre and platey andesite’. These lavas Caradoc Series Nematigraptus yracilis shales show low Lresicularity, flow lamination and vary in thick- Llandeilo Series Glytograptus teretiusculus shales ness from a few metres to 240 m. A lack of surface struc- - - - - Unconformity ------tures usually associated with viscous lava could indicate Upper Didymograptus murchisoni shales subsequc’nt erosion. No evidence has been observed for Main tuff group la\ra haLping been chilled or brecciated in water and it is Llanvirn Series Rhyolitic tuffs probable that eruption was subaerial. From the above DiRadnor Forest. The lower ground is usually mantled mesh (180 microns) fraction for chemical analysis. by glacial drift but the higher ground, underlain by the Drainage is not sufficiently well developed on the Builth Volcanic Series, is free of drift and covered with a higher ground underlain by the volcanic rocks to allow ef- thin layer of soil developed on a surface of periglacial shat- fective reconnaissance using stream samples alone. Con- tering. The ground is well drained and marked by an sequently two soil traverses across the volcanic outcrop unusual abundance of mole hills. and four soil grids covering basic lavas, known mineralisation, a dolerite body and an area of the main Previous mining activity tuff group where mineralisation was suspected, were There has been very little mining activity in the area sampled. Samples were collected using a 1.2 m long hand surveyed, which is in marked contrast to the intensity of auger except in some areas of shallow soil over the activity in the Shelve Ordovician inlier, some 35 km NE volcanic rocks where a spade was used. The samples ob- along strike, and in the Cwm Elan area 20 km to the west. tained using the auger usually represented C-horizon Lead is reported to have been worked near Llwynceubren material though it is possible that some collected on the Farm [SO 064 5901, east of Llandrindod Wells, from the lower ground failed to penetrate boulder clay which latter part of the 18th century until the mid 19th century o\cerlies shales. (Hall, 1971): At the south end of a line of workings there are three shallow pits and adjacent dumps containing Chemical analysis galena in association with abundant calcite but access to The analytical techniques employed for each sample type underground workings is now impossible. Exploration are shown in table 2. The analytical procedures have been pits extend for 3 km along a northerly trend from these developed by the Analytical Chemistry Group of the dumps. The workings are located in rocks of the main tuff British Geological Survey and have been described in group on the NW limb of an anticlinal flexure plunging to other reports in this series. the SW and acid lavas (keratophyres) are exposed to the south. To the SE is the NE plunging anticline which ex- Treatment of data poses the didymogruptus beds in the core. A trial level, 25 m long, is found in the stream known as Chemical data for drainage samples were investigated us- Murchison’s section on the north side of Gilwern Hill [SO ing simple single element statistical techniques with a 096 5891. The footwall shows slickenside striae and pro- view to the identification of sites showing anomalous con- bably represents a fault plane. Vein quartz is evident but centrations of elements of economic interest. Most no metallic minerals were observed. The locality is believ- elements showed distributions closer to log normal than ed to have been explored for lead (Hall, 1971) although normal so that statistical calculations were performed on some local farmers maintain gold was the target. Lead the log-transformed data. Exceptions were those elements was also believed to be the target of a shallow (2 m) trial for which most values fell close to or below the detection level located at the northern end of Llandegley Rocks [SO limit for the method employed (e.g. MO). 138 6211 b u t no evidence of mineralisation is to be seen Class intervals for the preparation of histograms were today. derived after the method of Le Peltier (1969) and cumulative frequency graphs were plotted on log- probability paper for all elements determined in each sample type. For straight line distributions the threshold MRP RECONNAISSANCE EXPLORATION for anomalously high values was usually chosen to include the top 2.5 per cent of the population, corresponding to Geochemical sampling the mean plus twice standard deviation for a simple As a first step in the survey samples were collected from lognormal distribution. In other cases class intervals were natural sediment traps in all major streams and related to the breaks in slope shown by the distributions or tributaries draining the Inlier. This resulted in 107 were chosen empirically. Where analytical determina- samples being obtained from a 65 km2 area, a density of tions showed most samples to contain levels less than the slightly under two samples per km2. At each site sediment detection limit (e.g. Ag in sediments, Sb in pans) or an was wet-screened through nylon mesh to minus 8 BSI element was present over a narrow compositional range mesh (2000 microns) and further wet-screened to give a (e.g. Zr in pans) distribution maps were not prepared.

4 Table 2 Analytical Techniques

Analytical Method

Sample AAS (nitric acid) AAS (perchloric acid) OES XRF - 100 mesh Cu, Pb, Zn, Ag, Co, Ni B, Sn, Cr, V, Mn, Zr, Ba, Fe As, MO sediment Soil Cu, Pb, Zn, Ag, Co, Ni Cu, Pb, Zn, Ni, Mn, Fe B, Sn, Cr, V, Mn, Zr, Ba, Fe As, MO Panned Ce, Ba, Sb, Sn, Pb, Zn, Cu concentrate Ca, Ni. Fe, Mn, Ti, MO

Table 3 Characteristics of element distributions Drainage reconnaissance

Boron (sieued sediment) Range 50 % ile 97.5%ile The maLlansantffraed-in-Elvel is unknown.

5 LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY BORON (sediment)

‘“~.=‘p’, , , , 1 10 50 90 99 Curnulative Frequency (x)

XClog) 74 ppm Range 23-156 ppm x+2s (log) 125 ppm

. < 99 ppm l > 99 C 125 ppm 0 > 125 ppm

GEOLOGICAL KEY

Upper Shales (Nemagraptus gracilis) (Glyptograptus teretiusculus) (Didymograptus murchisoni) Tuffs, volcanlcs and associated sediments

Lower Shales (Didymograptus bifidus) II Dolerite - - Fault ----- GeologIcal Boundary +Pb Former mine

KEY

= Roads A+B class, others

- - - Railway

1-- - Streams

rr] Towns

+ Small villages North

KIlometres ~ 0 1 2 3 I I ,“i l,l:l,l’, ,i’,‘,“,‘I’I I I 02 03 04 05 7 08 1 0 Miles 1 2

Figure 2 Distribution of boron in sieved stream sediment samples LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY CALCIUM (pan concentrate)

1 n=107

X(b3) 0.25 Z Range 0.070-2.524X 57+2s(log) 1.65 P/,

I . N ncx7 - V.“” m I a > 0.60-=1.6( 0 > 1.60%:

GEOLOGICAL KEY

Tuffs, volcanics and associated sediments

Lower Shales (Didymograptus blfidus)

I -I- - Fault ----- GeologIcal Boundary

_ Roads A+B class, others

- - - Railway

)---- Streams

Figure 3 Distribution of calcium in panned concentrate samples LLANDRIWDOD WELLS ORDOVlClAN INLIER DRAINAGE SURVEY TITANIUM (pan concentrate)

IC Ti (Z) 1.0

1 10 50 90 99 Cumulative Frequency (%) Wlog) 0.65 b Range 0.31 -3.34 % E+2s (log) 2.50 %

GEOLOGICAL KEY

Lower Shales (Didymograptus bifidus) 1 Dolerite

- - Fault ----- Gedogical Boundary *Pb Former mine

” I I I 02 03 04 05 06 I 08 1 0 Miles 1 2

Figure 4 Distribution of titanium in panned concentrate samples LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY

VANADIUM (sediment)

lo-

i 1’0 io 9-o $9 Cumulative Frequency (%) X(log) 90 ppm Range 37-135ppm sT+2s(lcg) 135ppm

. 130 ppm

GEOLOGICAL KEY

Upper Shales (Nemagraptus gracilis) (Glyptograptus teretiusculus) (Didymograptus murchisoni) Tuffs,, volcanics and associated sediments 7

Lower Shales (Didymograptus bifidus)

Dolerite

L - Fault ---- Geological Boundary

KEY

= Roads A+B class, others

--- Railway

)--- Streams h@$!$ Towns t + Small villages North

Kilometres 1 2 3 I I I I i 0 Miles 1 2

Figure 5 Distribution of vanadium in sieved stream sediment samples LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY

MANGANESE (sediment)

I n=107 I 1 10 50 90 99 Cumulath re Frequency (%I F(log) 1260 ppm Range 279- 11800 ppm R+2s(log) 6400 ppm c 2500 ppm > 2500 > 5000 ppn > 5000 ppm

GEOLOG ICAL KEY

02 03 04 05 06 7

Figure 6 Distribution of manganese in sieved stream sediment samples

Manganese (sieved sediment) Iron (sieved sediment) Most of the samples containing relatively high levels of Higher levels of iron in the sieved sediment samples are manganese are derived from the peaty areas underlain by predominantly associated with the outcrop of the Lower the Lower Shales. This association is probably a reflection Shales and may reflect the oxidation of pyrite associated of the preferential secondary environmental precipitation with these rocks and the secondary environmental of Mn in such terrain where drainage is relatively imped- precipitation of hydrous iron oxides in areas of impeded ed relative to the areas underlain by the volcanic rocks. drainage. Several of the corresponding concentrate samples are relatively enriched in Mn but to a much less extent which Iron (panned concentrate) suggests that the Mn is present dominantly as coatings to Samples with relatively high levels of iron occur mostly the grains and rock fragments rather than as discrete Mn- within the areas underlain by the Lower Shales and par- rich mineral grains. Apart from this Mn in concentrate ticularly over the Upper Shales and volcanic rocks to the shows some correlation with the samples richest in TiO,. east of Llandrindod Wells. In the latter area there is a

10 “nY”“Ib,nlY IIYLILll DRAINAGE SURVEY

-\ -5 ; L. .I:.:

Figure 7 Distribution of iron in sieved stream sediment samples

11 Figure 8 Distribution of iron in panned concentrate samples

positive correlation between iron and lead which may be a Copper (sieved sediment) reflection of a significant pyrite or oxidised pyrite content The range of copper in sieved sediments is low. There is in the samples especially as these are not rich in Ti as some suggestion that the outcrop of the Upper Shale se- would be expected if they reflected oxide derived from quence is marked by a slight enrichment of Cu. The basic igneous rocks. source of the four samples in the south of the area with the highest Cu contents is unclear but it is possible that they Cobalt (sieved sediment) are related to E-W faulting which is prominant in There is a strong positive correlation between Co and Mn southern part of the Inlier. None of the follow-up work levels in the sieved sediment, reflecting the strong associa- has provided evidence of copper mineralisation. tion of the two elements in the manganese oxide precipitate fraction of the samples. The two Co-rich Copper (panned concentrate) samples from south of Llandegley are richer in Co relative The range of copper in panned concentrate samples is to Mn than others and form part of a small anomalous much greater than that in the sieved sediments and this population the source of which is unclear. can be seen in the much steeper gradient of the concen-

12 I LLANDRINDOD WELl,S r\Dr\fi\,lP,AA, ,h,, ICP /= “““““,b,rrl” II’ILILrl DRAINAGE SURVEY

COBALT (sediment) I

Cumulative Frequency (%)

Range lo-105 ppm

. < 35 ppm l 2 35 < 50 ppm 0 C5Oppm

GEOLOGICAL KEY

Tuffs, volcanlcs and associated sediments

Lower Shales (Dldymograptus bifidus)

- Roads A+B class, others

53. - - - Railway

)---- Streams m Towns 52. t + Small villages North

51 Kilometres 0 1 2 3

I’““““1 ; 02 03 04 05 08 1 0 Miles 1

Figure 9 Distribution of cobalt in sieved stream sediment samples LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY

COPPER (sediment)

100 C&h 10 n=107

1 10 50 90 99 Cumulative Frequency (%) R(log) 16 wm Range 5-30 ppm x+2s (log) 30 wm . c 23 ppm 0 > 23430 ppm l >, 30 ppm

GEOLOGICAL KEY

Upper Shales (Nemagraptus gracilis) (Glyptograptus teretiusculus) (Didymograptus murchlsoni) Tuffs, volcanics and associated sediments sl

Lower Shales (Didymograptus bifidus)

l-l Dolerite L 1

- - Fault ----- Gedogical Boundary

KEY

= Roads A+B class, others

--- Railway

1--- Streams m Towns t + Small villages North

Kilometres 1 2 3 I I I 1 02 03 04 05 06 7 08 i 6 Miles 1 2

Figure 10 Distribution of copper in sieved stream sediment samples

14 COPPER (pan concentrate)

K(h) 9 ppm - 1 .f Range I-172ppm I x + 2swg) 9OPpm 77 &

GEOLOGICAL KEY

Figure 11 Distribution of copper in panned concentrate samples

Zinc (siwed sediment) tratc cumulative frequency plot. The agreement between__. . the two sample types is low with only three sites, all m the There is a general association between relatively high south of the area, showing coincident anomalies and no 1~\.cls of Zn in the stream sediments and the peaty ground sites with coincident anomalies of the highest class. It is de\reloped over the Lower Shale sequence. Thus there is a probable that some isolated concentrate anomalies which good correlation between Zn and Mn levels. Samples have no corresponding sieved sediment anomalies reflect from the upper part of the stream to the east of Howey metallic contamination, though this has not been checked ha\.? high levels of Zn relative to Mn and probably reflect by microscopic examination of the samples. This is also mineralisation and this may also apply to the sample from suggested from the association of Cu with relatively high NW of Yendre 1088 5741. le\~ls of Sn in some though not all of the anomalous con- centrate samples. Two Cu-rich samples without %uzc (panned conrenirate) associated tin anomalies are from sites on Silurian rocks The cumulative frequency plot of Zn in panned concen- close to the margin of the inlier though the source of thcst tratcs shows a sharp increase in gradient at around the is not clear. 350 ppm le\rcl which probably reflects the presence of a LLANDRI RIDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY

ZINC (sediment)

n=107 I I 1 I I 1 10 50 90 99 Cumulative Frequency (%) x (log) -180 ppm Range 90-450 ppm R+2s (log) 400 ppm . < 250 ppm 0 > 250 < 400 ppm 0 > 400 ppm

GEOLOGICAL KEY

Upper Shales (Nemagraptus gracilis) (Glyptograptus teretiusculus) (Didymograptus murchisoni) Tuffs,. volcanics and associated sediments

Lower Shales (Didymograptus bifidus)

Dolerite

Ll- - Fault ----- Gedogical Boundary IV‘Pb Former mine

Figure 12 Distribution of zinc in sieved stream sediment samples LlANDRlNDOD WELLS ORDOVICIAN INUER DRAINAGE’SURVEY ZI NC (pen concent rate)

1000

Zn l pv-0 100 i\ 1 n=107

%log) 148 ppm Range 51-785 ppm y+2s (log) 420 ppm

. < 230 ppm 0 2 230 C 350 ppm 0 > 350 ppm

GEOLOGICAL KEY

Upper Shales (Nemagraptus gracilis) (Glyptograptus teretiusculus) (Didymograptus murchisoni) Tuffs,, volcanics and Iassociated sediments

Lower Shales (Didymograptus bifidus)

IDolerite

- - Fault ----- Gedogical Boundary

KEY

= Roads A+B class, others - - - Ral lway 1--- Streams

ml Towns

+ Small villages North

Kilometres 0 1 2 3 Jn~~tI~itihIr~~ll 1 I I 1 1 0 1 2 L Miles

Figure 13 Distribution of zinc in panned concentrate samples

significant amount of sphalerite in an upper population. anomalies is uncertain but it is possible that samples The samples from the upper part of the stream to the east dcri\.ed from the Lower Shales to the north of Pendre [088 of Howey almost certainly reflect mineralisation especial- 574].[093 57 l] also reflect mineralisation in view of the ly as the corresponding sieved sediments are also relatilrely high le\.els of Zn in some of the corresponding anomalous. This is also supported by the follow-up sie\.ed sediment samples. drainage and bank sampling described below. The anomalous sample from south of C:wm-brith Rank [082 Strontiwn (/mumi concentrate) 6021 and the sample frorn further downstream also pro- Samples with relatively high levels of Sr occur mostly in bably reflects mineralisation, especiallv in \riew of the the southern part of the area. There is some agreement spectrum of anomalous elements present, though the between Sr and Ca, especially in samples derived from follow-up grab and and bank sampling failed to pro\,ide the Lower Shales and three samples are also relatively rich supporting evidence for this. Overall there is an indica- in Ka. I,a<-k of agreement between Sr and Ti-rich samples tion of‘ a general association between Zn anomalies and suggests that sphene is unlikely to be the main Sr-bearing the northern part of the outcrop of the main tuff‘ group. C:a-rich miner-al present in the samples. ‘I’he source of the other isolated concentrate zinc

17 LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY

STRONTIUM (pan concentrate)

L 1 10 50 90 99 Cumulative Frequency (%) Z(log) 150 ppm Range 40-470 ppm R+2s (log) 390 ppm . < 230 ppm 0 >, 230 < 380 ppm 0 >, 380 ppm

GEOLOGICAL KEY

Upper Shales (Nemagraptus gracilis) (Gly,otograptus teretiusculus) (Didymograptus murchisoni) Tuffs,, volcanics and associated sediments

Lower Shales (Didymograptus biiidus) 1 Dolerite

- - Fault ----- Gedoglcal Boundary

- - - Railway

)--- Streams

North

Figure 14 Distribution of strontium in panned concentrate samples

18 LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY ZIRCONIUM (sediment)

Zr 3pm 100

rk107 t- 1 -it+dd Jlative Frequency (“/,

X(b) 265 ppm Range 165-492 ppm Y+2s (log) 400 ppm

< 320ppm > 320 < 390 ppm > 390 ppm

GEOLOGICAL KEY

Upper Shales j.;,:<;.(Nemagraptus gracilis) . . .’ . . “’. . :‘i. . (Glyprograptus teretiusculus) m (Didymograptus murchisonil Tuffs,, volcanics and sl associated sediments ~.~.‘.‘.‘.‘.~ ~.~.‘_~.~,‘_~. Lower Shales ‘.~.‘.‘,‘.‘.~. ‘.~.‘.‘.‘.‘.‘. (Didymograptus bifidus) n. . . . El A- Fault ----- Gedogical Boundary C^_rn^” -:..,.

--- Railway

2--- Streams

North

Figure 15 Distribution of zirconium in sieved stream sediment samples

Zirconium (sieved sediment) Barium (siwed sediment) There is a relatively small range of Zr contents in the sic\.- Ka-rich samples are scattered throughout the area. Many ed sediments and a scatter of relatively high \.alues across of the samples are probably derived from mineralisation the whole area. as they coincide with concentrate anomalies and with enrichments in other metals.The relatively high level of 7’in (panned concentrate) Ka in the sieved sediment relative to the concentrate in Several relatively tin-rich samples occur, many, but not some samples e.g. at Cwm-brith Bank [083 6111 suggest all, of which are close to roads. Since levels of tin in com- that any baryte present is relatively fine-grained or more mon rocks and their consituent minerals and in basc- likely that other lighter minerals such as mica containing metal mineralisation are very low, tin anomalies serve as Ba may bc present. indicators of metallic contamination of human origin. Mineral examination of the concentrates is requir4 to Barium (panned concentrate) confirm that all other associated metal anomalies art’ also The range of Ba in panned concentrate is not high com- in contaminants. pared with many areas of Britain and it is therefore

19 DRDOVlClAN INLIER

,pm) TIN (pan concentrate)

n= 107 \ I \\ I/

, R+2s (log) 180 ppm L* l < 20 ppm 0 >20 <4Oppm > 40 I45 ppm

Figure 16 Distribution of tin in panned concentrate samples

20 LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY

BARIUM (sediment) 0,000 c

R(log) 400 wm Range 214-985 ppm E+2s (log) 825 ppm

. C 613ppm 0 >613<825ppm 0 > 825 ppm

GEOLOGICAL KEY

Upper Shales (Nemagraptus gracilis) (G!yptograptus teretiusculus) (Didymograptus murchisoni) Tuffs, volcanics and associated sediments

Lower Shales (Didymograptus bifidus) I Dolerite - - Fault ----- Gedogical Boundary

KEY

= Roads A+B class, others

--- RaIlway

)---- Streams 4 m Towns + Small villages North

KIlometres

Figure 17 Distribution of barium in sieved stream sediment samples

21 ORDOVICIAN INLIER DRAINAGE SURVEY

BARIUM (pan concentrate)

1 10 50 90 99 Cumalative Frequency (%)

. < 700 ppm l >/ 700 Cl270 ppm

0 >1270ppm

GEOLOGICAL KEY

Lower Shales (Didymograptus bifidus)

--- Railway

)---- Streams

North

UILTH WELL

Figure 18 Distribution of barium in panned concentrate samples

unlikely that substantial amounts of baryte are present. the Upper Shale sequence but the reason for this is Two of the most Ba-rich concentrates are not associated unclear. In contrast the Ce levels of some samples derived with anomalous levels in the corresponding sediment entirely or largely from the volcanic rocks is relatively samples and could be derived from minor relatively low. coarse-grained baryte mineralisation. Lead (sieued sediment) Cerium (panned concentrate) The cumulative frequency plot of Pb shows a sharp break The cerium content of many of the concentrate samples is in slope and a high gradient anomalous population com- very high, as in other similar samples derived from prising 3 samples. Of these the one from near Howey is sedimentary rocks of the Welsh Basin (e.g. Cooper and derived from the old lead workings and the one from others, 1984). The cerium in these areas occurs as nodules south of Cwm-brith Bank [082 6021, though con- of monazite or grain coatings which are thought to be of taminated with material of domestic origin, probably diagenetic origin (Cooper and others, 1983). In the Llan- reflects mineralisation. The Pb in the sample from SW of drindod Wells Ordovician inlier the maximum Ce levels Llansantffraed-in-Elvel [073 5341 may be derived from a are associated with the Glyptograptus teretiusculus zone of the mineralisation source in view of the lack of an associated

22 LLANDRINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY CERIUM (pan concentrate)

000 Ze 1v-n) 100

n=107

7 1 10 50 90 99 Cumulative Frequency (/)

X(log) 460 ppm Range 24- > 10,000 ppm y+2s (log) 10,000 ppm . c 1500 ppm l > 1500 10,000 ppm

GEOLOGICAL KEY

Upper Shales :__15: ; (Nemagraptus gracllis) xi3‘.::_ : :x (Glyptograptus teretiusculus) (Didymograptus murchisoni) Tuffs, volcanlcs and associated sediments I :. .,.,._.,.,.‘_‘_..,’ 1. Lower Shales ‘.‘.‘_‘_~,‘.’(Didymograptus blfidus) II~.~.‘.‘.‘,~.’

Dolerlte II

-I - Fault ----- Gedogtcal Boundary

51- KIlometres

I I f.llil,l;l~l~l;l,lI’;‘~( 0 1 I 2 3 1 7 08 1 0 wiles 1 2

Figure 19 Distribution of cerium in panned concentrate samples

23 LEAD (sediment)

Cumulative Frequency (%) alog) 41 ppm Range 30-710 ppm X*2s(log) 475 ppm . C 46 ppm 0 >46 <56 >56<90 ppm > 370 wm # screened data GEOLOGICAL KEY

(Didymograptus murchisoni) Tuffs,, volcanics and associated sediments

Lower Shales (Didymograptus bifidus)

1 Dolerite

A - Fault -- GeologIcal Boundary

- - - Railway

North

UILTH WELL

Figure 20 Distribution of lead in sieved stream sediment samples

high amplitude Sn anomaly, though mineralogical ex- tent several Pb anomalies from north of Builth Wells and amination of the heavy mineral concentrate is required to also adjacent to roads in the centre of the area are likely to confirm this. Slightly elevated levels of Pb are associated be derived from contamination. Anomalies likely to with several samples from the Howey Brook drainage reflect mineralisation within the main tuff group are pre- basin and from north of Builth Wells but the sources of sent south of Cwm-brith Bank 1082 6021 and east of these are unclear. Howey [091 5941 and [ 111 5941 and west of Llansantffraed-in -Elvel. Lead (panned concentrate) The cumulative frequency plot of Pb in panned concen- Synthesis of reconnaissance drainage data trate samples shows a very sharp increase in gradient at about 60 ppm which correlates with the appearance in the A synthesis of the highest amplitude metallic-element samples of a significant amount of a lead-rich phase. This anomalies is shown in Figure 22. Several of the corresponds to Pb minerals and to metallic contaminants anomalous samples also contain elevated levels of Sn which seem to be quite widespread in the vicinity of which represents a good contamination monitor in the human habitation. In view of their relatively high Sn con- absence of Sn mineralisation in the area. At some sites the

24 LLANORINDOD WELLS ORDOVICIAN INLIER DRAINAGE SURVEY

LEAD (pan concentrate)

1 10 50 90 99 Cumulative Frequency (XI E(b) 36 ppm Range 2-2491 ppm E+2s Wg) *300 ppm . < 60 ppm

l >60<160 ppm 0 >160ppm *Screened Data GEOLOGICAL KEY

Upper Shales (Nemagraptus gracil(s) (Glyptograptus teretmsculus) (Didymograptus murchisoni) Tuffs, volcanics and associated sediments 7

Lower Shales (Dldymograptus bifidus)

Dolerlte

- - Fault ----- Gedoglcal Boundary

KEY

- Roads A*B class, others

- - - Railway t %- Streams

ma Towns t Small villages + North

Kilometres 0 1 2 3 1 I I 1 1 L 02 03 05 06 7 08 1 0 Miles

Figure 21 Distribution of lead in panned concentrate samples

25 DRAINAGE SURVEY

IISTRIBUTION OF GEOCHEMICAL ANOMALIES

. Sample site (no anomaly)

Sample shows concentration of * indicated element(s) as follows:

pan concentrate -100 # sediment

Bap > 1270 ppm Bat > 825 ppm

Cup > 70 ppm Pbp > 300 ppm Pb., > 200 ppm

Sbp7 12ppm

Znp > 370 ppm Sn > 20 ppm (shown only at sites with anomalous concentration of another element.) GEOLOGICAL KEY

Lower Shales (Didymograptus bifidus)

Figure 22 Synthesis of distribution of base metal anomalies in drainage samples

presence of both contaminants and natural heavy FOLLOW-UP DRAINAGE SAMPLING minerals is suspected and detailed mineralogical examina- tion of the samples is required to confirm the presence of Follow-up drainage sampling was carried out in 11 ore minerals. The most prominant anomalies of Pb and streams where the presence of anomalies derived from Zn appear to be associated with the outcrop of the mineralisation was suspected, the location of which are volcanic unit, particularly in the north of the inlier. shown in Figure 23. The sampling consisted of the collec- Elsewhere there are more isolated anomalous sites, some tion of grab samples of stream sediment from natural of which are clearly contaminated and environmental fac- sediment traps at 100 or 50 m intervals up the stream. A tors are responsible at least in part for the elevated Zn minus 85 BSI mesh (< 180 micron) fraction of these levels in the sieved sediment samples from the central samples was obtained for chemical analysis. At the same Lower Shale area. time a sample of soil was collected from each bank at a depth of about 25 cm. The soil samples were dried, dissagregated and dry screened to give a minus 85 BSI mesh (< 180 micron) fraction for analysis.

26 Figure 23 Location of follow-up grab sampling drainage surveys

A clear confirmation of elevation of Pb and/or Zn levels RECONNAISSANCE SOIL SAMPLING was obtained in 4 of the 11 streams sampled and the most anomalous sites are plotted for these in Figure 24. The The reconnaissance drainage data were supplemented by highest amplitude and most persistent anomalies in the the chemical analysis of samples from 27 reconnaissance follow-up sampling occur in stream 1 which cuts across soil traverses, the locations of which are shown in Figures the outcrop of the volcanic unit. These are also confirmed 27 to 29. All soil samples were collected in Kraft paper in the lower part of the stream by anomalous levels of both bags from material obtained by soil auger from between Pb and Zn in the bank samples. Other sources are 30 and 60cm, oven dried, dissaggregated where possibly indicated by the Zn and Pb anomalies further necessary and dry screened to give a minus 85 BSI mesh upstream. Sampling of stream 7 suggests that the source (180 micrometres) fraction. Two long orientation of the anomaly is likely to be further upstream within the traverses (1 and 2) were sampled in the south of the area outcrop of the same volcanic unit which is also probably across most of the outcrop of the Ordovician rocks and 19 the case in stream 8. The source of the anomaly in stream elements (Cu, Pb, Zn, Co, Ni, Ag, As,B, V, Cr, Mn, Fe, 3 is likely to be very local in view of the close cor- Zr, Sn, Ba, U, Sr, Zr and MO) and loss on ignition (LOI) respondence between drainage and bank samples. determined by AAS after nitric acid attack, OES and

27 .Follow-up stream (ij N 100m . Stream sample site t 1 J

Pb in grab stream sediment Zn in grab stream sediment

l >70 ppm Pb l a380 ppm Zn t

Pb in bank samples Zn in bank samples

Follow-up stream @

Pb in grab stream sediment Pb and Zn in bank samples

l >70ppmPb l > 70 ppm Pb

Follow-up stream 0 Follow-up stream @ Pb in grab stream sediment Pb in grab stream sediment

l >70 ppm Pb

Figure 24 Distribution of Pb and Zn anomalies in follow-up drainage surveys

28 140

120

100

Zn PPm 80

60

40

20 120

100

80 Pb ppm 60

40

20

40

30 CU pm 20

10

0

50 -

40 -

As pm 30- 20 -

lo-

0

60- x335 ppm B

50 -

40 -

NI twm 30 -

20 -

10 -

V pm

3000 -

Wn ppm 2ooo 1000 -

6-'

=e % 4-

2-

30 -

zo-

10 -

0' 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10

[above

200 190 180 170 160 150 140 130 120 110 100 90 80 JO 60 50 40 30 20 10

Builth Wells Soil Line 1 1 :lO 000

Figure 25 Plot of Zn, Pb, Cu, As, Ni, V, Cr, Mn, Fe and Loss on ignition (LOI) levels in reconaissance soil line 1

29 ;oil anomalies in the Llandrindod Hall, Llwynceubren Farm area of old mining activity 1

: 5,000 307OOOE

- 260000 N

So11 Anomalies

0 >, 40 ppm Cu n 2 34 ppm As

2 160ppm Zn >, 130ppm Pb ---- >, 370 ppm Pb >, 600ppm Zn & 1% Pb

Figure 26 Distribution of Cu, As, Zn and Pb soil anomalies (Cu, As point, Zn, Pb contoured) around the area of old mining activity between Llandrindod Hall and Llwynceubren Farm

30 LLANDRINDOD WELLS

ORDOVICIAN INLIER

RECONNAISSANCE SOIL LINES

Distribution of Copper anomalies

. b 40ppm < 1OOppm Cu

0 3 1OOppm

KEY

Roads A+B class. others

RaIlway

Streams

Small wItages So11 lines I North Boundary of Maln turf group

Scale 1 50000

kllcmetres

Figure 27 Distribution of Cu anomalies in reconnaissance soil lines

XRF, as shown in Table 1, in soil samples taken 20 m are also associated with the Glyptograptus teretiusculus apart. A plot of elements with a significant range of con- shales. Line 2 also shows an association of As and very tents viz. Zn, Pb, Cu, As, Ni, V, Cr, Mn, Fe and LO1 low amplitude Pb anomalies with the volcanic belt but this values for line 1 is shown in figure 25 together with a is not plotted separately. geological section taken from the 1:25 000 geological Samples from the lines other than 1 and 2 were analys- map. The volcanic sequence is in general marked by low ed for a reduced list of elements. Lines 3 to 7 were analys- levels of Ni and higher levels of Mn, Fe and to some ex- ed for Cu, Pb, Zn, Ag, Co and Ni after nitric acid leach tent V than the sedimentary rocks. The plot shows some by AAS at the Grays Inn Road laboratories of BGS. For Pb and low amplitude As anomalies and a Cu anomaly to lines 8 to 27 Cu, Pb, Zn, Ni, Mn and Fe were determined be associated with the sequence of andesites, felsitic ag- by AAS after a perchloric acid leach by Mather Research glomerates etc overlying the main tuff group. In addition Limited of Rothbury, Northumbria. The majority of the there is also a low amplitude Pb and As anomaly lines cross all or part of the outcrop of the main tuff group associated with the top of the main tuff group and a roughly at right angles to the strike direction and spaced broader area of relatively high Fe, Mn and V associated between 250 and 500 m apart. The lines in the SW part of with the base of the unit. Isolated Pb and As anomalies the area also cross the spilitic andesites and other rock

31 LLANDRINDOD WELLS

ORDOVICIAN INLIER

RECONNAISSANCE SOIL LINES

Distribution of Zinc anomalies

I > 160 < 230ppm Zn

0 > 230<400ppm 7-n

+ 2 400ppm Zn

Figure 28 Distribution of Zn anomalies in reconnaissance soil lines

types overlying the main tuff group. Lines 3 to 7 crossed highest ground. Heavier soils are developed on the lower the area of old lead workings near Llwynceubren Farm in ground of the inlier over shale horizons and glacial till. the west of the area while lines 11 to 13 cross doleritic in- The location of Cu, As, Pb and Zn anomalies in the trusions in the Nematograptus gracilis shales and in the Glyp- area of the old lead workings near Llwynceubren Farm is tograptus teretiusculus shales in the NW of the inlier of Or- shown in Figure 26. The location of the traverses and the dovician rocks. cumulative frequency plots of the complete soil data set Soils are poorly developed on the volcanic rocks form- are shown in Figures 27 to 29. The large amplitude Pb ing the higher ground of most of the Ordovician inlier. anomalies, the levels of which were set empirically, are They typically have a light open texture with a thin located adjacent to the old excavations as are the highest (5-10cm) dark b rown humic horizon overlying a paler Zn anomalies and the low amplitude Cu anomalies. medium brown or buff silty soil containing rock There is also a halo of lower amplitude Pb anomalies fragments. Soil thicknesses vary from 15 cm to 1 m and around the workings and further Pb anomalies and one were probably developed on a shattered periglacial sur- relatively high Zn anomaly between 120 and 240 m to the face. The ease of working of these soils can be west of the northernmost excavation. Zn levels are much demonstrated bv the abundance of mole hills even on the lower in the most anomalous soils than the corresponding

32 -1LWYNCEUBA

MANGANESE ZINC / KEY * c* bpd,,,-- --A, __---_’ \ _- _.__ /- r- e-’ \ \ I ;‘+CHURCH \ LLANDRINDOD WELLS

ORDOVICIAN INLIER

RECONNAISSANCE SOIL LINES

Distribution of Lead anomalies

I l 80<160 ppm Pb

0 > 160c750ppm Pb

t l 750ppm Pb

KEY k 1.i - Roads A+B class, others

RaIlway

Streams cl Towns Small villages So11 lines I Boundary of Maln tuff group North

Scale 1 50000

Figure 29 Distribution of Pb anomalies in reconnaissance soil lines

amounts of Pb but a wider zone of anomalous Zn in soil is A loose group of low amplitude anomalies are present associated with the mineralisation, especially in the north towards the western margin of the main tuff group to the of the area sampled. Low amplitude Zn anomalies are the north of Builth Wells. The relatively high amplitude Cu only features on the lines sampled beyond the limits of the anomaly near the west end of line 2 1 [051 5351 is known mineralisation and the lack of Pb anomalies sug- associated with a zone of high Zn and Pb in soils. The gests that the vein Pb mineralisation does not extend location of the anomalous zone suggests that it may be beyond the limits of the workings. related to the roughly east-west trending Wern To fault The location of Cu, Zn and Pb anomalies on the recon- shown on the published geological map and which cuts the naissance soil lines are shown in Figures 27 to 29. Class sequence of volcanic rocks in the vicinity. intervals of anomalous levels were established from the There are relatively few Zn anomalies outside the nor- points of rapid gradient change on the cumulative fre- thern outcrop of the main tuff group which are discussed quency plots of the whole soil data set. Copper anomalies below. Most of the anomalies in the southern part of the are relatively few, tow amplitude and scattered except for area are isolated and scattered except for those around those associated with the old mining area between Llan- Wern To mentioned above. There are coincident Zn and drindod Hall and Llwynceubren Farm described above. Pb anomalies at each end of line 27 and in the middle of

33 Y”\ ""V v v ”

Figure 31 G eo 1ogy of northern part of main tuff group

line 14 but Pb is not associated with the higher amplitude DETAILED SOIL SAMPLING Zn anomaly towards the south end of line 14. The higher amplitude Pb anomalies are also largely Further detailed work was carried out around Pen Rhiw confined to the northern part of the area. There is a loose Frank in order to provide a better picture of the distribu- grouping of low amplitude anomalies to the west of the tion of anomalous Zn and Pb in soil and also because of outcrop of the main tuff group in the south of the area. the observation during the reconnaissance sampling that Though no anomalous zone can clearly be defined there is significant amounts of weathered pyrite occur in tuffs and a suggestion of an association of the anomalies with the lavas in the vicinity. The additional soil sampling com- mapped outcrop of the coarse feldspathic sandstones prised 4 lines at 200 and 400 m on either side of the shown on the 1:25 000 geological map. original line 10 along which samples were taken at 25 m Contoured plots of Zn, Pb, Fe and Mn in soils from the intervals. Contoured distributions of anomalous Zn and northern outcrop of the main tuff group are shown in Pb in the area are shown in Figure 32. The contour pat- Figure 30 and the mapped geological boundaries of the terns are complex but when considered in the light of the unit in Figure 31. Apart from the anomalies associated detailed photogeological interpretation of the area with the area of old mining activity at the west end of the described below they can be seen to represent anomalous outcrop there are broadly coincident diffuse Zn and Pb zones which broadly follow the local strike of the rocks. anomalies which follow the strike of the belt. Areas of The anomalous zones are marked by samples containing relative enrichment in Fe and Mn are also associated with Pb levels generally between 200 and 300 ppm but with the belt but a detailed correlation between these elements more isolated samples containing in excess of 1500 ppm and Zn or Pb is not apparent. The highest Zn and Pb Pb. Most of the area is also marked by low amplitude anomalies occur to the southeast of Pen Rhiw Frank with anomalous levels of Zn but the few samples containing levels reaching up to 2000 ppm Pb. more than 200 ppm Zn do not correlate with the lead anomalies.

35 KEY

0 Borehole LEAD (PPm) 1

,

\ CHARGEABILlTY(ms) APPARENT RESISTIVITY(S2m) I

Figure 32 Contoured plots of chargeability and apparent resistivity together with Pb and Zn in soil around Pen Rhiw Frank. Borehole locations also shown

Pits were dug down to about 50 cm at a few of the sites The outcrop surface of the acid lavas is commonly rusty containing the highest Pb levels and the profiles sampled and ochrous and when broken the rock is usually pale at three or more depths. Chemical analysis of these grey or buff. Rounded dark blue areas, 2 - 3 mm in samples for Pb, Zn, Mn and As did not reveal any diameter, represent either remnants of less altered rock or evidence of downward enrichment in the elements, the micro-xenoliths. Vesicles are frequent and usually contain levels being similar to those of the hand auger samples. quartz or secondary iron oxides or a combination of both. Samples with cuspate voids probably represent flow-top material. The groundmass of the rock is aphanitic and DETAILED GEOLOGY contains l-2 mm long feldspar microphenocrysts. Cut surfaces show the groundmass to have a speckled ap- To augment the geological information on the published pearance. The pale buff speckles (< 0.1 mm diameter) 1:25 000 map of the Llandindod Wells Ordovician Inlier a probably represent devitrification products such as photogeological interpretation of the area immediately sericite. Pyrite occurs as a fine dissemination in the around Pen Rhiw Frank was prepared using aerial matrix, in aggregates up to 2 mm in diameter and also as photographs at 1:6600 scale. The existence of three units vesicle fillings. Oxidation of the pyrite has produced a in the area was indicated. These were a) a well-bedded network of goethite throughout the rock and accounts for gently dipping sequence, b) a steeper-dipping unit which the general ochrous outcrop surface. forms crags on valley sides and c) hummocky ground In the tuff sequence poor exposure makes individual which shows no obvious orientation of features. In the beds difficult to identify. Outcrops are typically smooth field the above were found to correspond with the follow- and either buff or ochrous in colour and when hammered ing lithological units: a) acid lavas interbedded with tuffs, the rock often shatters into granules. Cut surfaces show b) pale grey well-indurated tuffs and c) acid lavas. A that most of the exposed rock is pumiceous lapilli tuff with geological map containing information obtained from the a buff devitrified matrix containing feldspar air photographs together with information taken from the microphenocrysts up to 1 mm in length. Vesicles are filled published geological map is shown in Figure 33. with quartz or a mixture of quartz and pyrite and there

36 are also large pyrite crystals, up to 1.5 cm in diameter with the fault deduced from the photogeological inter- which cut across the fabric of the rock. Lithic clasts of pretation it is parallel to this and marks the edge of the lav.a, usually vesicular, also occur. 25 m west of Pen Rhiw zone of high resistivity. Frank a hard dark grev tuff is exposed in an old excava- SP surveys were carried out on lines 00, 200NE and tion. It consists of pale blue-grey siliceous clasts, 1 - 2 mm 200SE but no anomalies were found. Similarly the total in diameter each with a rim of fine pyrite. The clasts ap- variation in magnetic field was less than 20 nT and accor- pear aligned and some are flattened with a long axis of up dingly the results have not been plotted. to 5 mm, indicating incipient welding. Pyrite occurs within the clasts, usually with a spheroidal form. DRILLING Structure The association of chargeability maxima with the highest The sequence forms the northwestern limb of a NE- levels of lead in soil within a sequence of acid lavas and plunging anticline, the axis of which runs along Howey tuffs containing significant amounts of visible pyrite in Brook and dips are broadly towards the north. The rocks very limited exposure was considered suitable for further near to the eroded anticlinal axis show steeper dips than investigation by drilling. Four holes were drilled to test those further down the limbs, presumably due to stress the source of the geochemical soil anomalies and to in- release as the core of the anticline in argillaceous rocks vestigate any association between this and the observed was eroded. The photogeological interpretation reveals geophysical anomalies. The location of the holes is given that the area around Pen Rhiw Frank is cut by a north- in Figures 32 and 33. trending fault and the eastern block defined by this fault is In the siting of boreholes a regional dip of 30 degrees to divided by a northwest-trending fault. Within the western the N was assumed from surface mapping. Borehole 1 was block dips are generally within 10 degrees east of north. sited to test the continuation at depth of the lead anomaly The northeastern block shows well defined and consistent 100 m SE across stike and to test the possible association photogeological dips 15 degrees west of north. Poor ex- of this anomaly with apparent resistivity and chargeability posure in the southeastern block allows few dips to be maxima, relatively displaced to the NW (i.e. towards the measured except in the case of the broadly northerly- borehole site). Borehole 2 was also sited to intersect the dipping basal tuff. source of soil lead anomalies across strike within the same fault block assuming them to be statabound. Borehole 3 was sited in the westerly fault block to intersect surface GEOPHYSICAL SURVEYS lead anomalies and a zone of relatively high chargeability down dip 200 m to the north. Borehole 4 was sited to help Geophysical surveys were carried out in the area around in interpeting the sequence of volcanic rocks intersected in Pen Rhiw Frank on lines generally either 175 or 200 m borehole 1. apart. IP/resistivity, VLF, SP and magnetic techniques The rock types encountered within the boreholes con- were employed. All IP/resistivity results were collected sist of acid/intermediate lavas and tuff. Tuffs were greatly with a dipole-dipole array of unit length 25 m and a predominant but a black vesicular lava occurred between separation of 50 m (N = 3) and this implies a nominal 5.55 and 6.0 m in borehole 1 and lava of the type termed depth of penetration of 50 m below surface. ‘keratophyre’ on the published 1:25 000 map, is predomi- Apparent resistivities are given in ohm-metres but due nant in the upper 20 m of borehole 2 and the upper 15 m to an instrument calibration error the true values are 1.66 of borehole 3. These ‘keratophyres’ when fresh are pale times greater. Apparent resistivity values have been grey homogeneous rocks with iron oxide-coated open ‘filtered’ using a running mean procedure to smooth out voids (cuspate or oval) or vesicles (usually quartz filled). variations caused by ‘thin’ resistive horizons. In this con- The open cuspate voids may represent cindery flow-top text ‘thin’ means a width comparable to the dipole length material. Dark grey cuspate patches (0.5 - 2.0 mm) and (25 m). The results are plotted in contour form in Figures mottling of pale grey within grey green probably 32 and 34. Values range from 60 ohm-metres to represents crystals or quartzose patches within a micro- 2000 ohm-metres and a recognisable zone of high crystalline groundmass of sericitised feldspars showing resistivity is discernible. weak flow alignment enclosing rare microphenocrysts ex- Background chargeabilities are very low, in the range 0 hibiting albite twinning. Fine pyrite is disseminated to 10 msecs but values rise to 25 - 30 msecs generally coin- throughout the groundmass and larger grains with broken cident with zones of high resistivity. The values in con- edges enclose the groundmass, possibly indicating tour form are superimposed on the apparent resistivity disintegration during flow. Pyrite crystals up to 2 cm map in Figure 34. The highest chargeability value of across also occur in veins. 29 msecs occurs at 250 m SE on line 200NE. In general the upper part of the tuff sequence in- Separate VLF base and traverse lines were surveyed tersected in the boreholes consists predominantly of because no suitable transmitting station exists for the pumiceous-lithic-lapilli tuffs containing horizons of traverse direction used for the IP surveys which was possibly reworked water-lain tuffs and probably cor- orientated nearly at right angles to the presumed strike of responds to the main tuff group recognised on the BGS the rocks. The station used was NAA (16.8 kHz) and the 1:25 000 special geological sheet of the area. The lower signal strength and null definition was only moderately part of the sequence comprises pumiceous lapilli tuff, a good. Cultural noise from fences etc was removed from well-lithified unit possibly equivalent to the basal rhyolitic the results by eye. Only one recognisable cross-over was tuff unit of the 1:25 000 map though compositionally it is apparent, on the eastern lines, and the position of this is not rhyolitic but andesitic. plotted in Figure 34. The clearest cross-overs, on lines The upper part of the tuff sequence contains a variety 350E and 490E suggest a conductive structure. Though of textural varieties ranging from tuff containing this might represent mineralisation it is more likely to fragments (0.5-2.0 mm) in a dark grey matrix to reflect a fault. Though this cross-over does not coincide pumiceous-lithic-lapilli tuff with clasts (up to 5 mm)

37 &j4> ...... :

...... ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~...... ::::::::::::::::::;:::::::::::::::::::::::::::::::::::::::::::::::kl. iiiiiiiiiiiiijiiiiiiiiiiiiiiiiiiiiiiiiii~*l* .~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~..l ., ...... if ...... ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. ,

.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. l l l I;l;l;l;l;r;l;r;l;l;r;l;I;l;l;l;l;l;l;l;:;:$ ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~,~.~ ~.~.~.~.~.~.~.~.~.~.~.~.‘.~.~.~.~.~.~.~.~.~.~.~.~.~.~ l l . ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.‘.~.~.~.~.~.~.~.~.~.~ l_ l. . . ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~ ...... i / .~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~ . l ...... * ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~ l ...... _[ ‘I ;- ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~ )? .x-) ...... ,. . ., ...>...... % ...... $ ,., ; :, : .. .~.-.‘.-.~.-.~.-.~.~.~.~.~.*.~.~.~.*.~.~.~.~ . . KEY

Road or track -w- Fault (photogeology)

Shaft or well;‘old working -L I L Topographic low

Borehole ---- Photogeological boundary Dip (from BGS Special Sheet SO 05/06 Llandrindod Wells) Valley (underlain by shales) Photogeological feature l l . _*.*l .l . Crags (comprising .:::/_.. . Steep dip (photogeology) l l . . 0. El. . l .- indurated tuff) Gentle dip (photogeology) Hummocky ground (underlain by acid law

[Figures refer to National Grid Square SO]

Figure 33 Photogeological map of area around Pen Rhiw Frank with horehole locations

38 KEY

0 Old pits, Mine workings

Survey lines

-111 VLF Xover

pa >lOOO Rm R8

>1500Qm

CHARGEABILITY

MC >20ms

>25ms

-- __F Fault

--- Geological boundary

- Crags n +BHl l --_) BH2 +~-> BH 3 @ mIneraked ~----t BH4 K = Keratophyre

I

CaO K,O 1

Figure 35 Triangular CaO-MgO-K20 plot of borehole samples in sequence down each hole

mainly of white or grey purple rhyolite or pumice developed lithification. Vesicles in the pale lapilli are filled (1 - 10 mm) in a devitrified dark grey of chloritic greenish with calcite and vesicles in both dark and pale lapilli show grey matrix. Thin sections demonstrate that lapilli and no signs of collapse. groundmass are both devitrified to sericite and secondary The core from all boreholes was split and samples taken iron minerals. Groundmass material is predominantly for chemical analysis at intervals based on lithology and subrounded ash fragments containing feldspar microliths differences in the degree of alteration of the pyrite and while polycrystalline quartz also occurs. Pyrite is variation in rock matrix. In some parts of the core the widespread in some parts of the core and two generations matrix appeared to have been hydrothermally altered and can be recognised. Within clasts pyrite usually shows these sections were split and a quarter of the entire length good crystal form while that found in the groundmass cuts taken for analysis. Other sections were selectively sampl- across the fabric of the rock and is clearly secondary in ed with a minimum of 20 cm to represent each metre of origin. In one specimen pyrite is concentrated in rims to core. The core showing little evidence of mineralisation clasts. Over much of the sequence pyrite is altered to was sampled in 5 m sections. secondary iron oxides and the rock is fissured and Each sample was passed through a jaw crusher and a permeable to such an extent that during drilling opera- 100 g split was ground to minus 300 BSI mesh (< 53 tions for most of the time there was no return of water. micrometres) using a Tema mill with tungsten carbide The pumiceous-lapilli tuff forming the lower part of the pot. Portions of the ground sample were taken for deter- sequence intersected in the boreholes contains subangular mination of Pb, Zn, Co and Ni by AAS after nitric- clasts (0.5 - 15.0 mm) of pale grey pumiceous and perchloric acid attack and of Cu, MO. Mn, Fe, Sr, Sb and subangular to cuspate dark grey pumiceous and Zr by XRF at the BGS laboratories in Grays Inn Road. microlithic clasts in a light grey matrix. Thin sections Additional determinations of MgO, SiO,, S, K20,Ca0, show that both lapilli and matrix are devitrified. Pyrite As and Rb were carried out by XRF at Midland Earth occurs as anastomosing growths sometimes associated Science Associates and 11 of the most mineralised and with 0.5 -2.0 mm wide calcite veinlets and thin sections pyrite-rich samples were analysed for Au at Caleb Brett show that it is always rimmed by and intergrown with Laboratories Limited, St Helens, Merseyside. calcite. Patches and veinlets of calcite are common The ranges of the elements determined in each of the throughout the rock and probably account for the well- boreholes is given in Table 4.

40 Pen Rhiw Frank Borehole 3 Location SO0843 6006 Depth (inclined) 128.30m Azimuth 183” Inclination 60” Depth %MgO %SiOz .%K*CI %CaO ZFe %Fe,O, %S ppm Pb wm Zn (m) 2 4 6 8 6 8 10 12 14 16 1 500 1000 200 400 F- 9xG-l I

clay ? gouge

5200 +

1

4020 +

40 -_-_-_-_-<

50 --_-_-_-_-

60 - _ _ _--7

_-----_

7fJ-_ - ~ _-IT

8(-J_-_-z---T

go ._------:

_-_------coarser

TO 128.3m

Key ""Y" """V Y """VV ""V Y" Vesicular lava (Keratophyre) Y"""" F "'4"""

Highly altered lava

Dacitic tuff

Andesitic tuff

Figure 36 Graphic log am plot of MgO, SiO,, K,O, CaO, FeZOx, S, Pb and Zn contents of Pen Rhiw Frank borhole 3

41 Table 4 Range of elements in boreholes

Borehole 1 Borehole 2 Borehole 3 Borehole 4

MgO% 2.45- 4.81 2.60- 5.42 0.62- 4.50 3.51- 4.95 SiO, % 48.72-61.33 54.90-60.73 51.60-69.96 53.74-58.53 S% 0.15- 0.97 0.28- 1.46 O.lO- 2.00 0.04- 0.48 K,O% 0.47- 2.49 0.80- 2.88 0.35- 4.87 0.19- 1.05 CaO% 1.22- 5.37 0.56- 2.49 0.19- 7.72 1.56- 2.17 Mn PPm 690- 1360 380- 1520 200- 1480 730- 1210 Fe&I, % 6.30- 9.69 6.81- 8.59 4.72- 15.84 7.58- 8.67 co PPm 15- 35 , 15- 50 lo- 30 20- 30 Ni ppm 25- 60 20- 80 20- 55 20- 30 cu PPm ll- 28 14- 26 12- 43 lo- 16 Zn PPm 87- 259 105- 135 9- 613 117- 172 As PPm

A comparison of MgO, K,O and CaO levels in each sent as galena in the main zone. Such galena would have sample from the various boreholes is given in the form of been converted into secondary minerals as result of ox- a triangular plot in Figure 35. This shows a great deal of idative alteration in a near surface weathering environ- similarity between samples from the lower parts of ment and possibly also within a previous episode of altera- boreholes 1, 2 and 3 which suggests that the same se- tion. Additional minor lead mineralisation is possibly quence is intersected in these boreholes. In contrast there associated with minor fault structures and brecciation bet- is a considerable variation in the samples from the upper ween 56 and 61 m down the hole. It is difficult to assertain part of these boreholes. This can be accounted for if there whether Zn as sphalerite was originally associated with is significant heterogeneity in the upper part of the the lead mineralisation as it would tend to be leached volcanic sequence both in terms of original chemistry and under the conditions that lead to the crystallisation of lead characteristics superimposed on this as a result of secondary minerals. Zn levels are also mildly anomalous hydrothermal alteration and mineralisation. The overall in borehole 3 between 86 and 91 m where vesicles are con- composition of the majority of the tuffaceous rocks ap- spicuous though no sphalerite has been identified in this pears to be andesitic although further chemical data on section. The composition of the ‘keratophyre’ lava the more immobile elements is required for a more detail- horizon overlying the mineralised zone is anomalous with ed classification. The graphic log of borehole 3 in Figure high levels of iron chiefly in the form of limonite of pro- 38 shows that significant variation in K and Ca levels oc- bable hydrothermal origin in some sections cur throughout the sequence but as no detailed Pyrite is conspicuous within much of the core and levels petrographic studies of the core have been carried out the of sulphur reach more than 2 per cent over 5 m. No cor- origin of this is not clear. The association of the highest relation between pyrite abundance and proximity to the level of K with the lead-rich sample from 56-61 m in base metal mineralisation can be established from the borehole 3 suggests that some potassium enrichment may core. In fact there is a greater abundance of sulphur in the accompany the base-metal mineralisation. Chemically the lowest part of borehole 3 and also the top of borehole 2, most distinct unit occurs in borehole 3 between 18.3 and apparently at two different parts of the sequence, but the 3 1.8 m. This unit is significantly higher in Si and lower in significance of this is not apparent. Eleven samples from Mg, Ca, Mn and Fe than the other rocks and is probably borehole 3, including some relatively rich in pyrite, have dacitic in overall composition. been analysed for gold by Caleb Brett Laboratories Ltd and all were found to contain less than the detection limit of 10 ppb. The low levels of As and Sb throughout the MINERALISATION core, with maxima of 20 and 11 ppm respectively, and similarly low levels in the drainage samples and soil In the borehole sections significant mineralisation is con- samples analysed also suggest that it is probably unlikely fined to a 6.4 m section ( 5.8 m estimated true thickness) that precious metal- bearing volcanogenic mineralisation of borehole 3 below an inclined depth of 14.9 m. Contents of significance is present within the area investigated. of Pb, Zn and S together with Mg, Si, K, Ca, Fe and K Assessment of the controls of the mineralisation is very are displayed together with a graphic log in Figure 36. difficult in view of the high degree of alteration and poor Though the mineralisation within the anomalous zone recovery of the main mineralised section. The site of consists almost entirely of lead, levels of Zn are also mildly mineralisation is at the interface between a localised anomalous over some of the section. The lead is present in dacitic tuff unit which is chemically distinct from the rest secondary minerals within a highly altered and soft clay- of the volcanic sequence and a lava horizon which is pro- rich section but in view of the presence of a small amount bably of local extent. The chemistry of the mineralised of galena within much less altered siliceous clasts at 25.36 sections is more consistent with the alteration of prex- and 26.50 m it is probable that the lead was originally pre- isiting volcanic rock than the introduction of material as

42 N Pen Rhiw Frank Borehole 3 S

Extent of surface 365.8 - lead anomaly (>80 ppm) BH3

Pumiceous lapilli tuff and acid extrusive rocks Single high Pb value (550. ppm)

243.8 - 128.3m Vertical Scale = Horizontal Scale (1 :2500) Topography enlarged from 1:25 000 OS map

Figure 37 Section to show interpretation of relationship between surface soil anomalies and Pen Rhiw Frank borehole 3

chemical sediment of exhalative origin. Since mildly the bulk of the tuffs. Bevins (1985) mentions the existence anomalous levels of lead occur over most of the in- of local pillowed flows and hyaloclastites and uses this tersected thickness of the lava, which is also marked by together with the nature of the associated volcaniclastic sections highly anomalous in iron, it is likely that the sediments both above and below the lavas to suggest that base-metal mineralisation is tied to the emplacement of for much of the time lava emplacement was in a sub- 1aL.a. Further difficulties of interpretation arise because of marine environment. The underlying tuffs which were in- the association of the mineralisation with rocks which tersected in the present survey are not considered by because of almost flat dip occupy topographically the Bevins ( 1985) however. Within a subaerial environment highest part of the area investigated. The drill holes sug- it is likely that mineralisation tied to a particularly erup- gest that the dips of strata around Pen Rhiw Frank are tive phase would be diffuse and irregular in distribution around 5 degrees to the north, which is significantly less unless concentrated by a particular well-developed linear than that portrayed on the published 1:25 000 geological or arcuate structure. Within a marine environment con- map (Earp, 1977). I n consequence none of the holes were centration of brines to give stratiform mineralisation is sited at suitable locations to intersect the stratabound possible though not always present, given the need for ex- mineralisation, which was the source of surface halative solutions of appropriate density and composition anomalies, at sufficient depth to be below the zone of sur- and a suitable environment for accumulation. Further face weathering. A possible relationship between surface work on the nature of the volcanic rocks and their en- anomalies and the mineralisation intersected in borehole vironment of deposition is required to establish whether 3 is shown in Figure 37. A further consequence of the sufficient concentrations of base-metal sulphide of poten- relatively flat strata is that a minor stratabound enrich- tial economic significance are at all likely within the area. ment in lead could give rise to an extensive overburden anomaly of the type observed in the Pen Rhiw Frank , area. CONCLUSION Of fundamental importance in an assessment of the significance of mineralisation within the Llandrindod Evidence of base metal mineralisation of probable strata- Wells Ordovician inlier is the accurate modelling of the bound type has been found in association with a lava-tuff environment of volcanism. In particular it is important to interface but the high intensity of alteration of the establish whether the volcanism took place within a sub- material has prevented an adequate assessment of its marine or subaerial environment. In general the volcanic nature and controls. Nevertheless it appears that the rocks can be seen to be situated within a sequence of mineralisation is associated with an intermediate lava marine rocks which though not generally deposited in a horizon that was probably errupted in a subaerial en- shallow water environment were close to the eastern vironment. The mineralisation appears to be of limited margin of the Welsh Basin. Though volcanism must have aerial extent and therefore of minor significance. The started within a submarine environment it is possible that volcanic association of the mineralisation suggests that a volcanic island was rapidly formed which continued to there is a possibility that base metal mineralisation of grow as volcanism continued. Furnes (1978) in a detailed greater significance could occur on the flanks of the study of the volcanic rocks of the area failed to find con- volcanic centre within a marine environment where clusive proof of either environment but concluded that the basins which could trap metal-rich brines might exist. evidence observable suggested that the lavas were However it is probable that this environment of subaerially erupted and reached a similar conclusion for equivalent age to the mineralised material is not exposed

43 in the area, being covered by younger rocks. Mineralisa- F~VRNESH. 1978. A comparative study of Caledonian tion could also be associated with the waning stages of volcanics in Wales and west Norway. D. Phil thesis Univ. volcanism in the area, in rocks younger than the main tuff Oxford (unpublished). group but there is no direct evidence of this from either H\I I, (;. W. 1971. Metal mines of Southern Wales. G. W. the drainage or overburden data. Hall, Westbury. ,JONES 0. ‘1‘. and l’t~r;r~ W. J. 1941. The Ordovician rocks of the Builth district. A prehminary account. Geol. Msg., 78, ACKNOWLEDGEMENTS 185- 191. - - 1946. The complex intrusion of Welfield, near Builth The British Geological Survey is indebted to landowners Wells, Radnorshire. Q. J, Geol. Sot. London, 102, 157- 188. within the survey area for their co-operation in permitting - - 1948. a) A multilayered dolerite complex of laccolithic access to collect geochemical samples, to undertake form near Llandrindod Wells, Radnorshire. b) The form and geophysical surveys and to carry out a programme of distribution of dolerite masses in the Builth-Llandrindod drilling. Dr D. Cooper is thanked for valuable comments inlier, Radnorshire. Q. J. Geol. Sot. London, 104, 43 - 98. on the contents of the report. Diagrams were prepared by - - 1949. An early Ordovician shoreline in Radnorshire, staff of the Keyworth Drawing Office under the supervi- near Builth Wells. Q. ,J. Geol Sot. London, 105, 65 - 99. sion of Mr R. J. Parnaby. Ko~EI.\\K B. I’., H~\~EI I s M. F., B~VIN\ 1~. E.. lio\cn R. A. and Dr NI\IEY I’. N. 1984. The Ordovician marginal basin of Wales. In Marginal Basin Geology: Volcanic and REFERENCES associated sedimentary and tectonic processes in modern and ancient marginal basins. Spec. Pub/. Geol. Sot. London, 16, BEVINS R. E. 1985. Pumpellyite-dominated metadomain 245- 269. alteration at Builth Wells, Wales-evidence for a fossil I,EI’EITIF:K (1. 1969. A simplified statistical treatment of submarine hydrothermal system? Min. Msg. , 49, 451- 456. geochemical data by graphical representation. Eton. Geol. , - KOKELAAR B. P. and DIJNKLEY P. N. 1984. Petrology 64, 538 - 550. and geochemistry of lower to middle Ordovician igneous rocks NICHOI \ (i. 1). 1958. Autometasomatism in the Lower in Wales: a volcanic arc to marginal basin transition. Proc. Spilites of the Builth Volcanic Series. Q. J. Geol. Sot London, Geol. Assoc. , 95, 337 - 347. 114, 137- 162.

EARP~J. R. 1977. Llandrindod Wells Ordovician inlier, SP\RIC\ Ft. S.. SICI~RIIS\ON H. and C\KF

44