Identification Of Deterioration Hazard Potential For Quarried Rock Slopes

The winning entry in last year’s Marston Award for best paper presented to a branch meeting

By Dr Dawn T. Nicholson

The 1999 Quarries Regulations require a proactive approach to risk management for quarried slopes. This means that potential hazards need to be identified in order that more detailed geotechnical assessments can be undertaken if necessary. The techniques used to effect this assessment are not prescribed though many approaches have been published. A difficulty arises, however, because many of the tried and tested hazard-assessment methods relate to quantifiable, large-scale failures acting along distinct discontinuity planes. But studies have shown that many of the greatest dangers from falling material do not involve large-scale failure along discontinuities, but rather small-scale deterioration. This is likely to be induced by superficial weathering activity combined with the damaging effects of blasting, rather than wholesale slope failure. This type of slope instability is much more difficult to quantify and predict. A further difficulty is that most hazard-assessment schemes do not allow any prediction of the likely nature of slope failure and yet this is a critical factor in determining its likely consequences.

This paper presents the results of a wide-ranging investigation of the deterioration of quarried rock slopes. A classification scheme is developed which categorizes a range of deterioration modes according to the frequency and velocity of movement, and the size of constituent material. The implications of each mode of deterioration are described in the context of hazard potential, identifying features and mitigation measures. The relationship between these deterioration modes and rock mass and material properties is examined, together with the slope morphology which results. Together, these characteristics can be used to assist in the identification of potential hazards from deterioration. They can also aid in the assessment of likely consequences of failure and allow better planning of appropriate remedial or protective measures.

Definitions and terminology

Deterioration is used as an umbrella term to describe the combined effects of physical and chemical weathering and some erosive agents. Deterioration occurs because when a new slope is excavated in a rock mass two fundamental changes occur. First, the act of excavation releases confining pressure on the rock mass, leading to expansive recovery. This is achieved by an increase in total void space and largely effected by fracture propagation and dilation. Secondly, excavation exposes a newly created rock mass surface to ambient environmental conditions, especially temperature and moisture fluctuations. As a result of these changes, that part of the rock mass which has been exposed by excavation is no longer in equilibrium with its surrounding internal and external environment. The natural response to this is for equilibrium to be re-established through progressive breakdown and erosion of the rock slope, and significant change can occur over the human timescale.

THE FIELD INVESTIGATION

The objectives of the field investigation were to:

• establish the extent of deterioration for quarried rock slopes in the UK
• characterize the nature of deterioration
• determine how deterioration hazards can be recognized.

Evidence of the consequences of deterioration were also sought, as were influencing and controlling factors.

Only quarried rock slopes exceeding 45? in gradient were considered. Soil, soil-like and unlithified materials were excluded except where they formed part of a slope substantially cut in rock. A wide range of rock types was represented, though in the UK there was inevitably an emphasis on rocks of sedimentary origin. Many of the slopes assessed were located in the north of England, there being a notable emphasis on locations in North and West Yorkshire and Cumbria.

The total number of slopes investigated was 116 based at 59 sites. Of the slopes investigated, 86% were situated in disused quarries, 4% were in active quarries and 7% were in semi-active quarries. A further 3% of slopes investigated were natural.

RESULTS OF THE FIELD INVESTIGATION

Deterioration: occurrence, consequence and mitigation

From the field investigation it is apparent that deterioration is widespread on quarried rock slopes in the UK. The nature of deterioration in terms of the volume and type of constituent material, the frequency of fall 'events' and the mechanisms involved, varies considerably. All forms of deterioration were represented, including the occasional fall of individual grains resulting from in-situ disintegration, semi-continuous ravelling of highly fractured rock masses, and isolated rockfalls involving large volumes of material. Deterioration occurred to some extent on every rock slope investigated, though in some cases was inconsequential.

Deterioration presented a greater risk at disused quarries where monitoring and remedial work was either absent or at best sporadic. The potential risk of injury to visitors was highlighted by specific references to such in field-geology guide books. Where disused quarries were in use as industrial estates, residential areas or caravan parks, the potential hazard to properties and structures from falling debris was obvious. Less-serious consequences of deterioration were also observed, including concealment of, and damage to, valuable geological exposures due to rock breakdown. It was notable that an inverse correlation between deterioration and adverse aesthetic impact exists. Slopes which had developed a weathering crust, for example, were the least visually intrusive. Conversely, freshly excavated slopes, especially where multiple blastholes were in evidence, often presented a negative aesthetic impact, though this varied with the nature and quality of the surrounding landscape.

In active quarries the treatment of problematic deterioration usually involved mechanical or hand-scaling, re-routing of haul roads or, more commonly, simple restriction of access. In disused quarries the most common approach to deterioration was to increase stand-off distance or, more rarely, use wire-mesh netting or another protective measure such as fencing. In some old disused quarries it was often unclear where ownership and responsibility lay, and if there is no legal public access, deterioration was allowed to occur unchecked.

Controls and influences on rock slope deterioration

Because there is such a wide range of potential influences and controls on deterioration it was not possible to discern any widespread geographical or lithological correlation. Nevertheless, some general trends were identified and these are considered below.

Rock mass and material properties

As expected, block release is more pronounced where there is both a high fracture intensity and wide fracture aperture. In rock masses which were highly fractured but where fracture aperture was very tight, or where blocks were tightly interlocked, slopes were much more stable. Deterioration was greater in highly fractured rock masses in weak material than in strong rock masses with an identical fracture network. The greatest deterioration occurred in slopes with both poor rock mass and material properties.

Although major discontinuity sets such as bedding planes and joints often determined the overall structure of a rock mass, it was the smaller, less persistent, often highly irregular and dense networks of fractures where most block release occurred. These small, non-persistent fractures are the type which are commonly formed from stress release, blasting, vegetation and weathering effects.

The type of deterioration varied in slopes with different rock mass and material properties. For example, in weakened materials with a medium-to-coarse granular texture (sandstone, gritstone, oolitic limestone etc) deterioration was dominated by in-situ breakdown, grain raveling and surface scaling. Block release was much more common in stronger, fractured rock masses.

Environmental, stress and engineering factors

Deterioration was enhanced where there was groundwater seepage, sometimes because of in-situ decomposition or disintegration associated with the presence of moisture, and sometimes because of water flow through the fracture network leading to block release. There was no clear relationship between deterioration and slope aspect, although north-facing slopes, or other slopes cast permanently in shade, retained much surface water. Therefore, they commonly had a widespread cover of moss, algae and general surface staining and thus material weathering was greater than on slopes which received sun.

Deterioration was greater on high-altitude exposed slopes than on their sheltered, low-altitude equivalents. There was a clear correlation between the presence of woody vegetation and intensely fractured zones. This does not necessarily imply any cause and effect since it might simply be that vegetation is opportunistic, establishing where fractures and weakened material provide conduits for root growth and where moisture supply is plentiful.

Older slopes, particularly those more than 100 years old, tended to be much more stable than their younger counterparts. To what extent this is a function of age and equilibrium, or of a fundamental difference in the excavation methods used, is unknown.

The field investigation revealed that a range of slope micro-landforms could be identified which were related to the deterioration processes acting. These were an important factor in allowing assessment of the deterioration mechanisms operating in each case.
Deterioration morphology can be sub-divided into three types:

1. Erosional landforms.
2. Depositional landforms
3. Process indicators.

Erosional landforms

Chutes: Chutes are quasi-channels down which loose material is transported. They are characterized by having debris piles at the foot, in some cases with a wide spread of debris in or on chute surfaces. Three types of chute were observed: Erosional chutes occur where rock material has been cut into by erosive agents, usually surface water runoff. They can also be formed from solution. At three slope locations it was noted that vertical channels produced by a pneumatic hammer were acting as micro-chutes, down which water and fines were being transported. Fracture chutes occur in fractures with a large aperture, usually due to fracture enlargement by wall breakdown. Structural chutes are not strictly an erosional landform; they are the product of the intersection of discontinuity planes inherent in the rock mass with the slope plane and might simply be a function of the excavation process and slope geometry. Nevertheless, they provide conduits for down-slope movement of material and commonly have a build-up of debris, soil and vegetation. They therefore function as chutes and are susceptible to surface erosion.

Overhangs: Structural overhangs can also result from the intersection of discontinuity planes with the slope plane and might be a function of excavation procedure and slope geometry. Erosional or composite overhangs can occur where more competent materials are undermined by the erosion of underlying weaker material or by basal undercutting illustration of in homogeneous materials. Overhangs can also result from indicators solution.

Cavities: Some features can be transitional between cavities and overhangs. Cavities of a wide range of sizes might form from solution. Localized and small-scale cavities often form along horizontal discontinuities such as bedding planes, probably representing the early stages of undermining. Man-made cavities such as mine adits also occur.

Macro landforms: A range of large-scale karstic forms such as buttresses and headwalls and collapse dolines have been reported in other studies on disused limestone quarries in
Derbyshire, and although these features were not observed in an advanced state of development in this investigation, incipient karstic forms were sometimes observed, notably in chalks and oolitic limestones. Incipient gullying was also observed in some very weak rock slopes, and large-scale palaeoweathering (limestone solution) features were exposed in others.

Surface scars: Scars formed when material has been removed from the slope were a very common sight in the field investigation. In some cases, scars were noticeable because the newly exposed rock was more weathered than the adjacent material, indicating that weathering had penetrated either through the material or along discontinuities, at least t o the depth of the scar. In other cases the reverse was true, in that the scar revealed fresh, unweathered material behind. This was often an indication that the material which had been removed was itself weathered. In yet further cases, scars took the form of hollows left when a large volume of material had been removed from the slope (eg a rockfall). Scars are an excellent means of locating the likely origin of debris found at the foot of slopes.

Depositional landforms

Depositional landforms are formed from the debris which results from rock slope deterioration and can be located at the foot of the slope or on the slope itself. They are a useful means of estimating the likely magnitude and frequency of deterioration mechanisms and can be conveniently divided into debris piles, scattered debris and isolated debris.

Debris piles: Debris piles are concentrations of debris which might be of a uniform constituent size or multiple sizes. Debris piles probably develop from one or a combination of two processes, either the fall of a large volume of material in a single event, and/or the semi-continuous fall of material from the same location on the slope. The gradient and lateral spread of debris is related to the nature (particularly the size and shape) of the constituent material, the velocity of movement and the trajectory angle of the material as it falls. Evidence from scars on the slope suggests that debris piles formed from single fall events tend to have more lateral spread and a shallower gradient, forming quasi-fans. Gradual accumulation of debris from ravelling produces more concentrated, steeper debris piles. This is particularly true for platey fragments from shale slopes and, to a lesser extent, for sand grains.

Scattered debris: Ravelling of more blocky material tends to produce an extensive scattering of material at the foot of the slope with some localized concentrations. Where these concentrations are absent, this indicates sporadic fall of material from a variety of locations and at different times. Constituent material size is often quite uniform, indicating a common control on block size.

Isolated debris: Some slopes had isolated rock fragments at the foot, indicating rare falls of material. Constituent material size was often very variable, indicating a variety of controls on block size.

Fracture infilling: A further form of depositional landform observed was the infilling of very-wide-aperture fractures, usually acting as fracture chutes. Infill material was highly variable and several types could be identified, including: fines resulting from the disintegration of fracture walls (effectively in-situ infilling); fines washed into fractures from soil or detrital material; blocks dropped into fractures (some subsequently being involved in block wedging); and mineral precipitates (eg veins, healed fractures etc).

Process indicators

Process indicators are features which give an indication of the cause of deterioration. These mostly relate to in-situ disintegration and decomposition but the roles of surface water flow and vegetation are also worthy of special mention.

Water flow: Since many mechanical and chemical weathering processes depend on a supply of moisture, evidence of surface or groundwater flow is usually good indirect evidence of actual or potential weathering activity. In numerous cases where water flow was indicated there was corresponding evidence of enhanced weathering and deterioration. In dry periods, a number of indicators can be used to identify locations of regular water flow, including dampness retained in infilling materials or the rock material (evident from a colour change compared with dry rock); surface staining; penetrative discolouration; surface growth of moss and algae; preferential development of honeycomb weathering; ripples and other flow structures in accumulations of fines; the presence of vegetation; flattened or 'draped' grass; discontinuities enlarged or rounded by dissolution or water flow; and individual laminae or thin beds picked out by water erosion.

Vegetation: As indicated above, vegetation is often an indicator of water flow. It might be associated with in-situ fragmentation, particularly along, and in the vicinity of, large fractures.

In-situ decomposition: In a range of rocks, including sandstone, oolitic limestone, chalk and basalt, exfoliation or 'onion-skin’ weathering was observed of the type that can lead ultimately to corestone development. In some cases it was related to the penetration of chemical weathering from joint boundaries, and in others to mechanical splitting around highly contorted sedimentary slump structures. Solution was commonly observed in limestone rocks, producing a range of erosional forms such as micro-solution pits, runners and surface rounding. Honeycomb weathering was common in Triassic sandstones. A further form of in-situ breakdown observed commonly in coarse sandstones and gritstones was the breakdown of intergranular cement such that the material could easily be crumbled.

In-situ disintegration: At the rock material scale this is usually manifest as general weakening, allowing material to be crumbled easily. Increases in surface porosity might also be evident. In-situ disintegration at the rock mass scale can be recognized largely by the fractures present. The increasing frequency of horizontal and sub-horizontal fractures near the surface, probably due to rebound, was a feature observed commonly, especially in massive and thickly bedded rocks. In contrast, stress-relief fractures induced by blasting were also ubiquitous but notably irregular with a relatively wide aperture. Intensely fractured zones were evident, some probably relating to blasthole locations and others occurring in association with vegetation. On a few occasions, notably in chalk and oolitic limestone, loose blocks lying on the ground at the foot of the slope were completely shattered in situ, probably the result of freeze-thaw weathering.

DETERIORATION MODES

It became evident early on in the field investigation that a wide range of deterioration modes was in operation. Determining which mode(s) was active in each case was considered to have an important bearing on assessment of the potential consequences of deterioration. A classification of deterioration modes is proposed based on field observations. More detailed descriptions are given in appendix B. Categories of deterioration mode are distinguished according to frequency of occurrence, relative velocity of movement and size of constituent material (fig. 4). Event magnitude can also be inferred from the deterioration mode. Each mode can be recognized on the ground by the products of deterioration (eg depositional landforms) and erosional landforms (eg overhangs and scars). Each deterioration mode is distinct in terms of its characteristics, identifying features, safety implications, and remedial and protective requirements.

Semi-continuous modes of deterioration

Five semi-continuous modes of deterioration were recognized:

Ravelling is the frequent and semi-continuous fall of material. Three size divisions are recognized: grain ravelling, relating to clay, silt, sand and fine gravel particles <20mm; stone ravelling, relating to coarse gravel and cobble-sized particles from 20mm to 200mm; and block ravelling relating to boulder-sized particles >200mm. Grain raveling can be transitional with wash erosion (see below).

Flaking is a form of raveling involving the frequent and semi-continuous fall of material with a distinctive platey form, as can occur in fissile rocks such as shales and slates.

Wash erosion involves the detachment and transport of fine material entrained in surface water run-off.

Solution involves the dissolution of soluble mineral grains and cementing material in aggressive acid solutions, including rainwater. When this process operates at the rock mass scale and begins to affect the character of the rock mass it can be described as karstification.

Flexural toppling is a slow, progressive deformation and sliding of layered strata, due to gravitational forces, upon removal of lateral constraint.

Sporadic modes of deterioration

Three sporadic modes of deterioration are recognized:

Fall is the occasional fall of individual fragments. The size divisions for ravelling are also recognized here to give grainfall, stonefall and blockfall.

Contour scaling is a special form of fall involving the infrequent exfoliation of thin layers of rock material formed parallel to the slope surface.

Slabfall and toppling involve occasional and infrequent falls of large, tabular slabs and rotation of large prismatic blocks. A typical 'a'-axis dimension of such slabs and blocks is 1m. Material of smaller dimensions which falls in the same way can be regarded as stonefall or blockfall as appropriate.

Isolated modes of deterioration

Three isolated modes of deterioration were recognized:

Rockfall is used here as a specific term to describe the fall of many blocks of varying sizes in a single, identifiable event, and might involve freefall, slide, bounce and roll or a combination of these.

Debris flow is the rapid transport of a mixture of coarse and fine particles in a partially saturated, grain-supported flow, and involves initial sliding and subsequent flow processes.

Rockslide is the rare, large-scale and rapid translational movement of rock, often along a distinct discontinuity plane. This mode is strictly outside the scope of deterioration but since smaller rockslides also occur, and since the mechanism is often largely weathering related, it is included for completeness.

Occurrence of deterioration modes

A record was made of the occurrence of the different modes of deterioration on rock slopes examined in the field investigation and a distinction was made between major modes of deterioration which dominated and those which appeared to be of a minor nature. On many slopes several modes co-existed, while on others there was a single, distinctive deterioration mode. On most slopes there was more than one minor mode of deterioration operating. These modes operated locally and/or had minimal deterioration effect. Occasionally, on slopes where deterioration was minimal, only minor modes were active.

The percentage frequency distribution of total occurrences of deterioration modes is given in figure 5. This chart shows that the fall of stone and block-sized particles, whether as ravelling or sporadic fall, is the most common mode of deterioration. Modes which relate to material properties (eg wash erosion, grain ravelling and scaling) are less important, while the large-scale modes of debris flow, rockslide and flexural toppling occur least commonly.

Figures 6a, b and c show the percentage frequency distribution of major deterioration modes for sedimentary, igneous and metamorphic slopes respectively. One of the most striking results is the spread of the distribution for each rock group. There would appear to be greater variability in deterioration mode for the sedimentary rocks, with increasing dominance of a few modes for the igneous and metamorphic rocks respectively. This contrast probably reflects the much greater range of mass and material properties represented by the sedimentary rock slopes investigated. Igneous rock masses, for example, are much more likely to be strong and either relatively structureless or widely jointed. Sedimentary rock slopes, on the other hand, might be weak or strong, structureless, blocky or, more commonly, strongly bedded. They are also more likely to be composite (layered), ie to have interbedded layers of contrasting lithology. These differences mean that a wider range of deterioration mechanisms is likely. Metamorphic rocks retain some of the characteristics of sedimentary strata and therefore ought to be transitional in the range of deterioration modes represented. In fact, the dominance of flaking probably reflects the large number of slightly metamorphosed turbidites examined which had developed a strong cleavage. Stone ravelling and rockfall occur commonly in all types of rock, reflecting the overriding importance of intense fracturing, probably related to blasting.

ROCK MASS TYPE

In the course of the field investigation it became apparent that deterioration modes were strongly related to three rock mass properties, namely rock type, fracture network and rock strength; and that distinctive rock mass types could be identified on this basis. These rock mass types are distinct from those identified in existing classifications. The primary distinguishing factors for each rock mass are the arrangement of fractures and the rock mass structure. There are three primary categories: massive rock masses are subdivided into weak and strong massive; layered rock masses are sub-divided into layered, fissile and composite; and blocky rock masses are subdivided into regular and irregular blocky. These descriptions refer to the pattern of fractures, not closed discontinuities. For example, a rock mass with well defined horizontal bedding but where boundaries between strata were closed, would be described as massive, not layered. Three subsidiary rock mass types are also recognized: intensely fractured zones; soluble rock masses; and composite rock masses. Data sheets relating to the different rock mass types are given in appendix A, providing details of characteristics, geological occurrence, associated deterioration modes and hazard implications. The frequency distribution of rock mass types observed in the field investigation is given in figure 7. The chart shows that layered and blocky rock masses are significantly more common than other types.

Relationship between deterioration modes and rock mass types

The relationship between deterioration modes and different rock mass types has been determined from detailed analysis of the data collected and is presented in summary form in figure 8.

This indicates that some modes of deterioration are closely associated with particular rock mass types, while others are independent. For example, weak, massive rock slopes are dominated by deterioration modes which focus on material breakdown, including grainfall and grain ravelling, while fissile rock masses are strongly dominated by flaking. On the other hand, scaling and wash erosion seem to occur in most types of rock mass.

ASSESSMENT OF DETERIORATION POTENTIAL

Assessment of rock slope deterioration hazard potential using the guidance resulting from this study is non-quantitative and can be undertaken quickly, with minimal equipment and by non-specialists. It is not intended to be a replacement for specialist advice, but to precede and complement it. It is inevitable that specialists are not available onsite at all times and therefore day-to-day monitoring must be undertaken by non-specialist quarry personnel, or in the case of disused quarries, by other non-specialists such as landowners, educators and conservationists.

It is first necessary to consult the data pertaining to rock mass type (appendix A) and to determine which of the seven primary types of rock mass most closely represents the rock slope under consideration. The relevant data sheet will indicate the nature of deterioration (ie deterioration mode(s)) likely to be associated with that rock mass. The detailed data sheets for deterioration mode (appendix B) can then be consulted on that basis. These provide information which will allow a qualitative assessment of the likelihood that the particular mode of deterioration will occur, including an indication of its likely magnitude, frequency and, hence, consequences. The data provided also give detailed guidance on treatment measures for situations where it is necessary for costly stabilization to be undertaken, of the sort which is more commonly associated with engineered slopes. However, this is increasingly becoming necessary for quarry after-uses which might involve close-proximity access to the slope by members of the public.
For slopes which are already displaying signs of deterioration, further information can be obtained from the descriptions of deterioration morphology and process indicators.

Some of the remedial treatments referred to in the deterioration mode data might be inappropriate in certain situations. For instance, shotcrete application and the use of underpinning might compromise the need to minimize visual impact. Remedial treatments which might be more appropriate for quarried rock slopes include restoration blasting, vegetation establishment, backfilling or other substantial regrading.

CONCULSIONS

A detailed field investigation of quarried rock slopes in the UK has led to the development of classifications of deterioration landforms, deterioration modes and the rock mass types with which they are associated. These provide a simple, qualitative system by which non-specialists can make preliminary assessments of deterioration hazard potential and highlight areas requiring further detailed analysis. This is a particularly valuable development in light of the fact that many of the slope hazards arising from quarrying relate to small-scale, shallow breakdown rather than deep-seated failure. The assessment method presented herein is particularly suited to disused faces within active quarries or to disused quarries which are being developed for alternative uses.

Dr Dawn T. Nicholson, the author of this paper, which is presented here in an edited form, received the Institute of Quarrying's 2001 Marston Award for best paper presented to a branch.