Response to Consultation on HS2 Phase 2b November 2016,
submitted in February 2017 by Carolyn Warburton, geologist and
geotechnical engineer, re specific geotechnical and geological
challenges facing the HS2 route north of Crewe that in 2022 which
are still relevant to the 2022 HS2 High Speed Rail (Crewe –
Manchester) Hybrid Bill.
Document made available to Mid Cheshire Against HS2 with Carolyn
Warburton’s permission to make the report available to others
Are you responding on behalf of an organisation or group?
Route between Middlewich and Pickmere: Change the route over 26km in the Middlewich-Northwich
area to avoid brining and gas storage infrastructure and to minimise the risk of subsidence due to
underlying geological conditions. This builds on intelligence received during the 2013 consultation.
Question 2 Do you support the proposal to change the alignment and raise the route through the
Cheshire salt plains? Please indicate whether or not you support the proposal together with your
1. My Background ……………………………………………………………………………………………………….. 1
2. I do not support the proposal ……………………………………………………………………………………. 2
3. The information on subsidence ‘evaluated’ for determining the preferred option is
inadequate …………………………………………………………………………………………………………………….. 2
4. Examples of rail, and high speed rail, developments for which an appreciation of existing
subsidence was available and an engineering solution adopted – which still failed. ……………….. 3
5. Salt extraction and potential pillar failure in the Winsford Mine ……………………………………. 5
6. Controlled pumping and potential pillar failure in the Holford Mine………………………………. 7
7. Subsidence from Wild Brine Pumping ………………………………………………………………………… 9
8. The unconsolidated deposits affect cavity formation and present their own subsidence
challenges ……………………………………………………………………………………………………………………. 10
9. Future stabilisation of salt cavities …………………………………………………………………………… 11
10. Future Monitoring ………………………………………………………………………………………………. 13
11. Remedial works used for cavities………………………………………………………………………….. 14
12. Distribution of Risks and Responsibilities ………………………………………………………………. 15
1. My Background
I am a Fellow of the Geological Society, have a degree in Geology an MSc in Geotechnical
Engineering and an MA in Land Use Planning. I worked as a geologist, and subsequently
geotechnical engineer for British Coal opencast for 18 years. I moved to the Department of the
Environment when they were seeking a geotechnical engineer in Land Use Planning, and later to
Somerset County Council and the Welsh Assembly to work in Minerals Planning. I looked at the HS2
proposals out of interest, but the concerns these raised have prompted me to reply to the
2. I do not support the proposal
I note the proposed route refinement and why the Secretary of State is minded to make this change.
Whilst the amendment to avoid the greatest concentrations of existing brining and gas-storage
infrastructure is sensible for strategic and construction reasons, the proposed amended route will
not minimise the risk of subsidence. Predictions of subsidence require accurate data. These data do
not exist for HS2.
Whilst is recognised that the HS2 report Salt Related Ground Stability (SRGS) is a preliminary report,
it has been used to make a major recommendation/decision. Once this becomes the ‘preferred
option’, it becomes increasingly difficult to change, no matter what problems become apparent.
Alternative routes should be considered. Salt Related Ground Stability, identifies in 8.9 Further
Work to provide for ‘a more meaningful appraisal’. It is essential that this further investigation
should take place before the preferred option is selected. Once that step is taken, the argument
Not only will a failure to understand the ground conditions result in inadequate design,
inappropriate methods of construction, cost and time overruns, and impact on safety – I do not
believe it is technically feasible to meet the required tolerances. Ultimately, it will be essential to
assess these risks by a comprehensive site-investigation, but in selecting a preferred option, HS2 Ltd
must make itself aware of the stability problems to be addressed and of the limitations of
Savings on site-investigation is a false economy. Any saving is short-lived. Engineering consultants
will say that they can meet the specifications and tolerances, but of course companies want to do
the works – it could well cost £100 million for the ground engineering in this section. And if the
works fail, it’s a court case – until the taxpayer meets the cost.
These engineering costs have to be an issue; the subsidence risks multiply the costs of the
infrastructure. ‘As the money paid in fares will not repay the costs of building and maintaining the
lines and HS2 is unlikely to cover even the operating costs the government has to cover the shortfall
out of taxes’1. Every extra million pounds to fill a hole in the ground is taxpayer’s money. The private
sector is not attracted by the high infrastructure and maintenance costs, and the low income from
operations. ‘This is because the numbers simply do not stack up when it comes to return on
3. The information on subsidence ‘evaluated’ for determining the preferred option is
Network Rail is clear:
‘17.1 General Requirements Geotechnical design shall be based on the findings of geotechnical
investigations – these comprise a gathering of information about the Site and a ground investigation
(which itself comprises a desk study, field investigations and laboratory testing). The scale and cost
of the investigations should vary with (inter alia) the types and characteristics of the ground; the
availability and reliability of existing geotechnical information about the Site, and the size, type and
cost of the structure being designed.’
This information is not available; this is a red flag.
Whilst I am not arguing that a full site-investigation is required at this stage, the first stage, the desk
study addressed in SRGS, needs to be robust. And it isn’t.
A preferred option has been determined, and for some stretches amended, without a full
appreciation of the ground conditions that will be encountered. Subsidence owing to the extraction
of minerals or the abstraction of fluids has serious effects on structures and services. In considering
the route of HS2, it is essential to consider whether subsidence might occur; how significant will it
be; when; what form will it take. Importantly, is it both feasible and viable to provide engineering
solutions which will limit the effects to within acceptable tolerances and preclude catastrophic
4. Examples of rail, and high speed rail, developments for which an appreciation of existing
subsidence was available and an engineering solution adopted – which still failed.
As an indication of why ground conditions should be one of the first considerations in determining
the route, consider, firstly the Madrid-Barcelona High Speed Rail.
A critical report by the consulting firm KPMG, 20042, pointed to ‘a lack of in-depth studies and overhasty
execution of works’ as the most important reasons for the problems that dogged construction
of the Madrid–Barcelona high-speed AVE line. The evidence for potential problems was there3:
the Madrid–Barcelona high-speed line shows dissolution-induced subsidence and
compaction of infills and embankments.
The pre-existing conventional railway and the high speed line corridors traverse large
sinkholes previously documented in geomorphological maps.
sinkholes frequently and recurrently affect the railway network in this sector of the Ebro
In 1991, a sinkhole caused the derailment of a freight train on the old Madrid–Barcelona
The occurrence of sinkholes beneath this railway is relatively frequent and some particularly
unstable stretches have speed limitations. ‘A minimum probability of sinkhole occurrence of 45
collapse sinkholes/km2/year has been estimated in this sector of the valley. During the construction
of the AVE tunnel near Barcelona, a number of nearby buildings suffered damage from a sinkhole
that appeared near a commuter rail station, damaging one of its platforms.’2
The occurrence and activity of sinkholes in carbonate and evaporite karst terrains is one of the main
causes of subsidence-related damage and accidents on conventional railways (Guerrero et al.,
An extensive literature review lists a number of subsidence-events affecting railways, the following
highlighted section is taken from this5.
The presented literature review reveals that railways are particularly sensitive to ground subsidence,
especially collapse sinkholes. The presence or occurrence of a sinkhole in a railway track may imply
2 commissioned by ADIF (Administrador de Infraestructuras Ferroviarias) at the behest of the Ministry for Public Works
(Ministerio de Fomento) on 23 June 2004,
3 Railway deformation detected by DInSAR over active sinkholes in the Ebro Valley evaporite karst , Spain et al Nat. Hazards
Earth Syst. Sci. 2013
4 Spain J. P. Galve1 , C. Castañeda2 , and F. Gutiérrez3 1Departamento de Geodinámica, Universidad de Granada, Published
in Discuss.: 2 November 2015
5 Engineering Geology 102 (2008) 62–73 A sinkhole susceptibility zonation based on paleokarst analysis along a stretch of
the Madrid–Barcelona high-speed railway built over gypsum- and salt-bearing evaporites (NE Spain) Jesús Guerrero,
Francisco Gutiérrez, Jaime Bonachea, Pedro Lucha
costly maintenance works, service interruptions and in the worst situations, the derailment of the
train accompanied with losses in human lives.
Examples of railway lines damaged by settlements caused by other processes than dissolutioninduced
subsidence, pertinent to Cheshire, include:
Railway lines in the Yangtze Delta, including the magnetic suspension highspeed railway of
Shangai, have suffered from differential subsidence due to aquitard consolidation induced
by over-exploitation of multilayered aquifers (Yin et al., 2006);
Mining subsidence involving in some cases the rapid occurrence of collapse depressions has
damaged the Pennsylvania and New York lines in the United States (Klugh, 2001) and the
Furness railway in northwest England.
In 1974, a sinkhole 90 m across and around 70,000 m3 in volume formed catastrophically
beneath the Missouri–Pacific railroad in Hutchinson (Texas). The crater resulted from the
upward propagation (stoping) of a cavity created by solution mining of a Permian salt unit
located at a depth of about 130 m (Walters, 1978; Johnson, 1998);
Sagging caused by peat consolidation has locally deformed the Rannoch Moor railroad track
in Scotland (Waltham, 1989).
The Manchester-Crewe line in northwest England requires frequent maintenance works
(raising the track, reballasting) along an active subsidence depression whose activity was
accelerated by uncontrolled extraction of brine from the karstified rockhead of a halitebearing
Triassic formation (wild brining) (Waltham, 1989; Cooper, 2001).
In the Harz Mountains of Germany, near Bad Sachsa, dissolution of gypsum has caused a
railway to settle more than 40 cm. A laser monitoring system was installed across the
subsiding section, which has a slow speed limit.
In 1997, a sinkhole around 30 m in diameter formed beneath the track of the Burlington
Northern Santa Fe railway in Missouri, causing the derailment of 22 railroad cars, fire from
spilled fuel and injuries to personnel. This sinkhole, induced by dewatering to allow mining
of limestone in a nearby quarry, has undergone reactivations in 1999, 2000 and 2002 despite
intensive grouting programs (Abkemeier and Stephenson, 2003).
In Allentown, Pennsylvania, a sinkhole triggered by a rapid decline in the water table
provoked by the interception of a phreatic conduit in a limestone quarry undermined a pier
of a railway bridge over the Bushkill Creek (Waltham et al., 2005).
Other sinkhole damages on railways in the United Sates have also been reported in
Pennsylvania (Perlow, 2003), Birmingham, Alabama (Newton, 1984) and Deland, Florida. In
Russia, cover collapse sinkholes formed beneath the Moscow–Nizhny Novgorod line have
caused the train to derail two times (Tolmachev et al., 1999).
In China, sinkholes induced by water withdrawal have caused service interruptions to the
Beijing– Jiulong railway line (Lijun, 1997). According to Chenguoliang (1990), as many as 7
sinkholes, all of them located close to pumping wells, have occurred beneath the Litang–
Zhanjiang railway line. This author also reports that a 32 km long stretch of the Guiyang–
Kunming line, between Ganhaizi and Yangtianchong cities, is frequently affected by collapse
In Fakou city area (China), where more than 1950 new sinkholes have been inventoried,
subsidence has led to the abandonment of a total length of 45 km of railways (Yu, 1994;
Gongyu and Wanfang, 1999).
Wenhui (1990), in a compilation of “geological calamities” in Chinese railways, indicate that
the detrimental effects generated by the numerous sinkholes reported in the railway
network include the derailment of two trains, more than 2000 hours of traffic suspensions
and maintenance and repair costs of thousands of millions of RMB (1 renminbi≈0.13 US$).
Van Hecke et al. (2003) indicate that high-speed railways are those that allow trains to run at speeds
higher than 250 km/h. It is considered that 15 mm is the maximum settlement tolerable for a train
running at that velocity. With higher vertical deflection the train may derail and with lower values of
vertical strain, velocity reductions may be necessary to diminish vibrations (Miura et al., 1998;
Woldringh and New, 1999).
At least three high-speed railways have been adversely affected by subsidence:
In Sweden, the speed of the train that connects Goteborg and Malmö had to be reduced due
to vertical deflections of 6 mm related to the presence of soft soils in the ground (Woldringh
and New, 1999);
According to the newspaper Taiwan News, in 2006 a high-speed train derailed two times
during the testing period due to differential settlements related to consolidation of
aquitards in the Yangtze Delta;
In 1993, a TGV train derailed at a speed of 294 km/h at Haute Picardie (France) due to a
sinkhole formed over a buried trench from World War One and triggered by heavy rain (Cui
et al., 1995; Brabie, 2005).
Some of the factors that affect the probability of a train-derailment due to sinkholes in karst areas
The length of the railway built on ground susceptible to sinkhole activity;
The sinkhole hazard including the probability of occurrence of sinkholes and their severity,
which refers to the size of the sinkholes and the velocity at which the settlement occurs
(subsidence rate) (Gutiérrez et al., 2008b, d). Obviously, collapse sinkholes that form in a
sudden way without showing any previous noticeable sign of instability are the most
The vulnerability of the infrastructure, which depends largely on the incorporation of
sinkhole protection measures in its design and the installation of monitoring devices linked
to a warning system capable of detecting subtle deformations and anticipating the
formation of collapse sinkholes (Tolmachev et al., 1999; Guerrero et al., 2004);
The frequency of the trains and their velocity. The latter determines their tolerance to
settlements and the breaking distance (Miura et al., 1998; Esveld, 2001).
5. Salt extraction and potential pillar failure in the Winsford Mine
A number of the conditions described above are known to affect HS2. Some of the basic issues are
described below, as the SRGS appears to accept assurances of residual caverns being permanently
stable, without evidence and without question.
The proposed HS2 route crosses the Winsford mine. The pillars carry the load of the roof, and are
subject to compression. Pillars will deteriorate with time due to both chemical and physical
processes which are difficult to characterise or predict. The voids in the collapsed rock above may or
may not reach the surface – with a sudden catastrophic failure.
Factors affecting their integrity are6:
height/width ratio of the individual pillars
depth below rockhead
height/width ratio of the cavity
maximum stress at the edge of pillars
pillars in the centre of workings are subjected to greater stress than those at the periphery
6 Bell, Engineering in Rock Masses, 1992
Faulting can result in a line of weakened pillars
Progressive spalling from the pillar-edges reduces the strength until eventually collapse
The failure of one pillar can have a domino effect
Whilst there are empirical formulae for pillar strength trial and error was the principal
method. The ‘design’ was often for maximum mineral recovery to provide stability only in
the short-term, in steady-state conditions
The water environment
The rock mass in terms of rock strength, structure and discontinuities
‘Empirical rules of thumb like the 10x seam thickness rule often cited and referred to in CIRIA SP32,
Healy and Head (1984), are open to misinterpretation and abuse. There are various scenarios in
which the rule should be used with caution’. 7
The Winsford mine has an areal extraction rate of 68-75%, based on work from the 1950’s. The
resultant pillars appear to be worryingly small. There is Planning Permission for extraction of a
second seam – we do not know how potential subsidence will be compounded by this extraction.
The potential for pillar failure, together with the subsidence of brecciated ground (see later), during
the construction and operation of HS2 is high for a number of reasons (I have cited examples from
coal mining, but the issues are the same):
Increased load – static from embankments and construction, and dynamic under
construction and from the trains as they pass
Decreased load – under cuttings. (This too increases spalling of pillars, see High Lane
opencast site and colliery, North Staffordshire.)
Abstraction of water from outside the immediate area. (Dewatering of Patent Shaft
opencast site resulted in subsidence of mine workings up to 10km away. Dewatering of
Sudeley opencast site may have destabilised bridge abutments a kilometre away)
Recharge of fresh water with renewed solution of salt
Change in water pathways. Compacted fill, or grouting under, for example, road corridors,
provides an effective dam to underground water. The former resulted the severe flooding of
land adjacent to opencast sites near Telford and hazardous subsidence to the West of
Merthyr followed construction of the A470
Salt Related Ground Stability:
4.2.2 states that mining has intercepted brine-filled fractures, but ‘assuming that the
potential inrush has been brought under control, these would not be expected to be a cause
8.4.1 ‘The operator indicates that the ground movements are minimal…’The monitoring data
must be assessed by HS2 Ltd
‘ it is expected that the mine will be maintained beyond the term of its planning consents for
extraction’. This is inadequate.
8.4.2 ‘the long term proposals, say on abandonment, have not been identified.’
‘Significant water ingress into the mine could radically change the apparent current stable
8.4.3 ‘Access to settlement monitoring data, in particular once the second seam mining is
underway’. It is not clear how the timing of mining and HS2 construction might relate?
7 Civil Engineering & Mining Related Geohazards – A Clients Guide To The Regulatory Process Leigh Sharpe, The Coal
‘Daily monitoring along the track could reduce the risk rating from high to medium or low’.
The speed at which collapses occur does not make this a safe option.
6. Controlled pumping and potential pillar failure in the Holford Mine
This statement comes from the Cheshire Replacement Minerals Plan, adopted 1999 and carried
‘This method involves the creation of a solution of salt and cavities in the salt beds remote from
zones where natural brine may be found. The shape and nature of the cavities formed are
monitored by sonar equipment to ensure surface ground stability. Once cavities have reached the
designed shape and size they are normally left full of saturated brine and sealed. This method of
extraction introduced to Cheshire in the 1930’s is considered to be permanently stable and there has
been no subsidence in the controlled brinefields. The present controlled brinefields extend to nearly
20km2 and extraction takes place in cavities ranging from 120m to nearly 250m below the surface.’
The caverns reach a height of 200m, a diameter of circa 150m and are developed initially on a grid
some 200m apart.
In controlled pumping, water is introduced under pressure to raise the brine to the surface,
dissolving salt from the cavity walls until maximum size and shape are attained. In the past it was
then capped off and left full of saturated brine. (Current practice often involves use of the caverns
for gas storage: the revised route does not traverse the gas storage caverns).
Whilst the pillars are not, reportedly, as small as in the Winsford mine, it is of interest to compare
extraction rates with the up-to-date analysis requirement of 275m between wellheads in the Keuper
Gas Storage Project in the Holford Brinefield – with caverns of 80-100m diameter, an estimate of not
much more than 10% extraction by area8.
The Holford Gas Storage Project9
‘Comparison Design Salt Mining vs Gas Caverns:
Typical Salt Mining: Distance between caverns: 200 m Cavern Diameter: 140 m Pillar to
Diameter ratio: 0.4 *
Gas Storage Caverns: Distance between caverns: 300 m Cavern Diameter: 100 m Pillar to
Diameter ratio: 2’
*Note the abstraction at Holford has a cavern diameter of 150m, with commensurate
reduction of pillar size.
Cheshire Brine Subsidence Compensation Board (CBSCB)10 says the cavities are designed to remain
permanently stable. They would. It is interesting to note the debate in Parliament when the CBSCB
was set up. ‘Will this system of controlled brine pumping stop subsidence forever? There are
different views about this. Professor Edgar Morton, who is one of the greatest experts in the country
on this subject, has given as his considered view that one day the salt pillars will give way, and that
the resultant collapse will mean a disaster comparable only with that of an earthquake’.11
It is not clear how, or whether, operations are monitored; they are outside the remit of the CBSCB
and do not appear to be required by Planning Condition. (‘No documents found’ on the website12).
The Minerals Plan (1999) says ‘The shape and nature of the cavities formed are monitored by sonar
equipment to ensure surface ground stability’. These data should be made available and assessed.
It is quite possible that the sonar monitoring constituted a final survey to verify the size of the
caverns at the time of completion; whilst subsequent sonar measurements take place at Preesall, it
is not known what monitoring takes place elsewhere, other than for gas storage..
Salt Related Ground Stability Report
4.2.2 ‘expects monitoring of movements may form part of the planning consent’; these data should
be made available and assessed. Presumably the permissions for the storage of ethylene do require
8.5.2 ‘Monitoring is reported to indicate very small magnitudes of subsidence but no detailed results
have been provided to verify the report…There is a report of significant ground movement having
8.5.3 ‘Inovyn have reported an historic incident where a blow out of brine occurred some 200m
removed from the causative cavern operation…’
‘A number of the caverns in Holford appear to have minimal intact rocksalt strata between adjacent
‘Confirmation of a long-term maintenance and monitoring programme could reduce the risk to
moderate’ It is difficult to know what form of maintenance could counteract the deterioration of the
caverns, and again, monitoring may not be sufficiently timely. The longer term problem, should the
track be affected, is how to remediate the problem. The line would be closed.
Sonar must be supplemented by surface subsidence measurements, to identify the movements
which are a consequence of the convergence of the salt rock mass. Monitoring of the pillars by
geophysics is critical.
Abandoning salt cavities to achieve a stable solution is a challenge, best met, but at a cost, by
backfilling. The industry prefers to fill sealed cavities with water or brine, but this has serious
questions for the long-term stability (Lux, 2009, identifies eleven), with the potential formation of
sink-holes and subsidence. The following must be considered:
Salt moves slowly under large pressure differences. In a cavern in salt, the pressure is lower
than the pressure from the overlying rocks, and the salt squeezes towards the void. For gas
storage, there needs to be a sufficient distance between storage caverns, generally in the
order of 300-400m13.
Assuming long-term (permanent?) tight borehole sealing exists, creep of rock-salt over time
will result in an increase in fluid pressure. At these near lithostatic pressures, rock salt is no
longer impermeable, and micro-fractures connect and are propagated, with consequent
rock fall. This reduces the strength, and increases porosity and permeability.
In gas-storage terms, it has been considered permissible to allow limited weakening and spalling,
provided third party protection is fulfilled. It is critical for HS2 Ltd to understand:
13 Evans and Chadwick, 2009 Underground gas storage: worldwide experiences and future development in the UK and
Europe. Geological Society of London Special Publication 313
how does the infiltration zone develop?
how much time does it take to build up to hydraulic breakthrough?
do the mechanical properties change?
are there consequences to the construction of the well bore seal?
HS2 Ltd must take into account:
the loads to be expected during construction and operation, as well as after abandonment
the specific rock mass structure and properties
Interbedded salt and marl, faulted and with cavities
The depths of the workings by no means preclude subsidence, particularly given the significant
depths of unconsolidated deposits. Walters (1977) records sink holes over brine fields in America.
In 1974 a 90m diameter sink hole developed over 3-days, resulting in railroad tracks being
suspended over a subsidence hole. A 90m diameter sink-hole developed over a 12 hour period at
Barton County, when unsaturated brine came into contact with a salt face 300m below the surface.
Again, in America, Landes and Piper (1972) describe sink holes of 60m and 135m developing over
salt workings at 400m depth.
7. Subsidence from Wild Brine Pumping
Subsidence from Wild Brine Pumping continues; it will be exacerbated by the shift from equilibrium
conditions engendered by the construction of HS2
In Cheshire, wild brine pumping resulted in subsidence, normally linked to brine runs where fresh
water entered the system14. The exact area from which salt was extracted was not known or
recorded. Subsidence took place up to 8km from the pumping centres, demonstrating the solution
of salt well outside the abstraction permitted areas. Collapse above the brine runs gave rise to
flashes – linear water-filled hollows – up to 10m deep and 90m wide. The flanks may show tension
scars, usually with displacement of less than a metre, but along which further movements may
occur. In the area resulting from erosion of the Top Bed, subsidence of deep troughs was a sudden
Subsidence did not necessarily occur immediately after the removal of salt; cavities collapsed years
after pumping ceased.
‘Following the cessation of the majority of the natural brine pumping in the 1970’s, the Board
established a network of precise monitoring stations in areas where ongoing subsidence had
previously been observed, and bi-annual surveys were carried out in the transition period from
natural to controlled brine pumping between circa 1980 and 1983’. This implies that monitoring
stopped in 1983. It is important also to recognise:
The tightly drawn areas selected by CBSCB, selected to minimise compensation
The emphasis on retrospective compensation rather than prediction and prevention
‘The Brine Board is not connected with the industry of salt extraction and cannot therefore
supply details of current natural or controlled brine pumping activities or any proposals for
the future development’.16
14 Bell, Genske and Stacey, 2000; Evans and Chadwick, 2009
15 Subsidence: Occurrence, Prediction and Control, D.J. Reddish, B.N. Whittaker
CBSCB argued that following the cessation of wild brine pumping, to explain continuing damage,
consideration needs to be given to potential alternative causes of ground movement including
natural seasonal movements, natural brine flows, and other forms of water abstraction for example,
along with structural defects in the buildings and adverse ground conditions.’ In other words,
subsidence is continuing.
Salt Related Ground Stability Report
5.1.9 ‘monitoring was introduced during the mid 1950’s’; these data should be made
available and assessed.
8.7.3 ‘There is no firm evidence that these features are actively subsiding’ Have
investigations been made? The monitoring and evaluation described later in the paragraph
should be completed before the route is determined.
8.7.4 ‘Whilst the specific magnitude of natural/wild brine pumping movements have not
been quantified, it is believed that these are of comparatively low orders of magnitude’. This
statement must be substantiated, and related to the tolerance of HS2.
8. The unconsolidated deposits affect cavity formation and present their own subsidence
The geology of this section of HS2 is briefly described in SRGS. I would emphasise:
The upper surface of the rock salt, referred to as the ‘wet rockhead’ is overlain by a zone,
which can vary between 65 and 160 m thick, of a dissolution breccia of mudstone. This
comprises a residual unit of broken and collapsed material that typically is brecciated and
permeable. It is derived from non-soluble sediments that originally overlay or were
components of the halite formations. Some brine filled cavities are present, especially
towards the base of the breccia. This is far in excess of everyday geotechnical problems.
The collapsed unit should be regarded as a superficial deposit as the material is no longer in
situ. Brecciation in the overlying beds results from wetting, and salt permeation by capillary
action, with eventual collapse into the cavity. This then leads to more general surface
subsidence, with continuing consolidation of the brecciated beds contributing to settlement
over the long-term.
The drift is complicated by a network of deeply incised channels, many of which appear to
have undulating depths and must have functioned as subglacial meltwater drainage
systems.17 These thick variable sequences of porous sediment form paths exploited by
In Pleistocene and Flandrian Natural Rock Salt Subsidence at Arclid Green,18 the respective
roles of Triassic bedrock halites, collapsed strata, periglacial alluvial sands and multiple
glaciation in determining the local stratigraphy and allied landforms are discussed. It is
concluded that natural rock salt dissolution is the principal process influencing the
superficial deposits and geomorphology of the study-area and that this process has been
active over hundreds of thousand years.
The variable drift has its own problems, and makes remediation of salt workings a more difficult and
A derailment in Canada resulted from a sudden punching failure through peat. Prior to the
failure, settlements at a rate of several centimetres per year could, over the eight years
before the derailment, have reached a value of 30 cm. Railway inspection procedures and
17 Howell & Jenkins, 1976, Some aspects of the subsidence in the rock salt districts of Cheshire, England, Int.
Assoc. Hydrological Science 121
18 Pleistocene and Flandrian Natural Rock Salt Subsidence at Arclid Green, Sandbach, Cheshire Peter Worsley
technologies, based on surface observations, are not able to detect the impending risk of
Trains operated by CN in Canada derailed along main lines 57 times in 2014.20
The construction of the electrified double tracks railway project at the Northern Peninsular
Malaysia commenced in 200721. The subsoils encountered vary from soft alluvium deposit to
dense residual soil, with thickness varying from 15m to 20m. ‘The geometrical tolerance of
railway tracks is stringent, especially for train with high operating speeds of 180km/hour.
The performance requirements include differential settlement of not more than10 mm over
a chord length of 10m and settlement of not more than 25mm within 6 months after
The EDTP project was awarded on a design and build basis for a lump sum amount of
RM12.48bn in July 2008. The contract was awarded with a precondition that any extra cost
related to the project was to be borne by the main contractor. The project cost has now
increased to RM16.5bn
9. Future stabilisation of salt cavities
Since the 1970s, many disused mine workings have been successfully stabilised by filling with grout
based on pulverized-fuel ash and Portland cement. However:
Drilling and grouting conventionally applies to relatively shallow workings.
Steel casing will need to be drilled to rockhead or to full hole depth if necessary to
prevent the ingress of fresh water
Drilling is likely to need brine-water flush, as under Northwich (see below)
While open mine workings can be filled with pfa/cement or sand/cement paste, sand
or gravel, grout for collapsed workings requires the use of thixotropic or polymeric
The creation of perimeter barriers using granular materials is seldom effective, see
grouting of the Heads of the Valleys Road where grouting was abandoned at a number
My experience does not lead me to believe that grouting at these depths, through these depths of
broken ground, without introducing fresh water and causing new dissolution, can be done to meet
the necessary specifications. But it will be possible to spend a great amount of money on failing.
It is salutary to consider the stabilisation of old salt workings under Northwich.
This would not meet the tolerances required for High Speed Rail.
The cessation of monitoring after one-year does not allow the effectiveness of the works
to be assessed22
Http://Www.Tsb.Gc.Ca/Eng/Rapports-Reports/Rail/2004/R04q0040/R04q0040.Pdf19 Railway Investigation Report
R04q0040 Main-Track Derailment Canadian National Train U-781-21-17 Mile 3.87, Lévis Subdivision Saint-Henri-De-Lévis,
Quebec 17 August 2004
21 Ground Treatment Design for 200km Electrified Double Tracks Railway Project at Northern Peninsular Malaysia Tan
Yean-Chin G&P Geotechnics Sdn Bhd, Malaysia Lee Peir-Tien G&P Geotechnics Sdn Bhd, Malaysia
22 In response to a FoI request to the council How often is subsidence in Northwich measured?
Official monitoring stopped in 2008 in accordance with the contract and planning permission. This was only one year after
the works were completed.
The interaction with other, unmitigated, causes of subsidence means that harm cannot be
attributed to any specific cause – see Requirements for Future Developments below
The work cost £38,000,000 with English Partnerships paying £33 million and the former
Vale Royal Borough Council £5 million. (FOI)
Ove Arup reports are highlighted below, emphasis is mine23
Since their abandonment in the 1920’s, salt mines beneath Northwich had deteriorated to such
an extent that the possibility of their collapse presented a threat to life and property.
The project involved the infilling of four salt mines to ensure the future stability of the town
centre. It was the largest project of its kind in the world, with 780,500m3 of void filled at depths
of 90m below ground. 32ha of surface land has been released for regeneration.
This stabilisation was undertaken between 2005 and 2007 under the auspices of the Land
In addition to the four so-called ‘Bottom Bed’ mines there are other processes (caused by a
complex combination of brine extraction and mining in the “Top Bed” above these mines that
are resulting in ongoing settlement in some parts of Northwich both within and outside the mine
area. The infilling works have not, and were not intended to, address these processes.
Post infilling verification has shown that the mines have been successfully stabilised with grout,
with an allowable void of 50mm.
The degree of infilling and the strength and stiffness of the placed grout are sufficient to prevent
any pillar failure or other collapse of the mines. A small amount of long term movement related
to the compression of the grout under loading from the mine roof, and long term creep of the
mine roof and pillars, is expected. The horizontal tensile ground strains associated with these
movements will be less than 0.15%.
The location and extent of any settlement is not easy to predict and any design work should
consider the following;
lateral brine flow
vertical brine flow
Requirements for Future Developments should consider the following, either singly or in
horizontal ground strain relating to the treated mines
settlement in the strata overlying the treated mines due to other processes
settlement relating to any untreated mines
presence of treated and untreated mine shafts
use of appropriate foundation design to overcome settlement generated by the above
treatment of Top Bed mines
the strength and likely performance of made ground and superficial deposits
Further site specific data collection in the form of site investigation, ground level
monitoring and ground water level monitoring may be required.
The Monitoring Regime24
23 Wrekin Construction Company Ltd Northwich Mines Stabilisation Programme Development Manual
J:\116000\116965-00\4 INTERNAL PROJECT DATA\4-05 REPORTS\PROJECT MANUAL\ISSUE FEBRUARY 08.DOC
Page 2 Ove Arup & Partners Ltd Issue 28 February 2008
24 Stabilisation of abandoned salt mines in North West England T.G. Brooks , N.J. O’riordan , J.F. Bird , R. Stirling & D.
Whilst the pillar stress analysis suggested that the mine pillars had an adequate factor of safety and,
therefore, were not at risk of imminent failure, settlement monitoring throughout the area indicated
that significant movements were taking place at localised spots above the mines. These movements
could have been associated with continued or accelerated settlement of the roof of the mines or,
alternatively, with completely independent processes taking place at wet rockhead or even in the
A dense network of surface levelling points was established… The evidence and
interpretation strongly suggested that surface settlement patterns are almost
exclusively related to processes occurring in the strata above the mines, i.e. at wet
rockhead and in superficial soils.
Magnetic extensometers and rod extensometers measured the relative settlement of
the various strata and any long-term movement above the roof of the mines.
Water levels and salinity profiling were monitored
HS2 Ltd should obtain and evaluate these monitoring results.
10. Future Monitoring
Monitoring must be undertaken to gain an understanding of ground behaviour. Repeat observations
over a meaningful period, linked to weather and extraction activities, are essential. Planners are
now required to take into account significant changes in sea-level related to climate change.
Suggested approaches include:
Light Detection and Ranging (lidar)
Interferometric synthetic aperture radar (insar)
Continuous GPS (CGPS) measurements
Global positioning system (GPS) surveying
Aquifer-system compaction by extensometers
Some of the above may be available for back-analysis, whilst the desk-study must collect and analyse
existing monitoring from all available sources.
The potential value of microgravity surveys is made clear below:
‘In the village of Marston, the Trent and Mersey Canal crosses several abandoned salt mine workings
and previously subsiding areas, the canal being breached by a catastrophic subsidence event in
1953. This canal section is the focus of a long-term monitoring study by conventional geotechnical
topographic and microgravity surveys. Results of 20 years of topographic time-lapse surveys indicate
specific areas of local subsidence that could not be predicted by available site and mine
abandonment plan and shaft data. Subsidence has subsequently necessitated four phases of
temporary canal bank remediation. Ten years of microgravity time-lapse data have recorded major
deepening negative anomalies in specific sections that correlate with topographic data. Intrusive
investigations have confirmed a void at the major anomaly.’
‘When field data are carefully acquired, processed, and combined with available site data in
standard numerical modelling, the results can be used to quantitatively inform interested parties
with not only the location and extents of suspected problematic subsidence, but also the potential
subsidence rates and timing of potential ground collapses’.25
11. Remedial works used for cavities
The difficulty for this stretch of HS2 is primarily the high risk of subsidence associated with salt
workings, but this is compounded by the broken ground and reworked glacial drift above. Multiple
treatments will be required at any one location to provide the degree of certainty needed for high
speed rail. Any answers must ensure safety, manageable maintenance costs and acceptable
remedial solutions should the track be closed.
As discussed above, grouting seems unlikely to meet those requirements, and, where relevant,
would not resolve the problems of the overlying broken ground and drift.
Particular issues arise at bridges and portals. Piles are often the solution, but in these circumstances
would be pushing well beyond current technology, for both physical limitations and corrosion.
The deepest piles to date were installed at the Pentominium; 250 piles for the tower to a
depth of up to a record 65m to support approx 3500 m2 of footprint area. (US$400 million
construction contract, currently stalled because of finance)
The amount of corrosion for a steel pipe pile can be categorized; for a pile embedded in a
non aggressive and natural soil, 0.015 mm per side per year can be assumed from the British
Steel Piling Handbook. Euro-code 3 now specifies various corrosion rates based on the
nature or soil conditions and pipe pile exposure
The M7426 A legacy of Glasgow’s industrial past was a series of mine workings, where seams
up to 2m thick are prevalent at depths of 30m to 40m below ground. £18.9M was spent on
these seams with a cement/PFA grout mixture, drilling more than 800km of bores and
installing 57,000t of grout. All the structures on site have piled foundations; 20,000m3 of
bored piles and 128,000m of precast piles were installed.
This engineering work, impressive in itself, is an order of magnitude shallower, and a fraction
of the areas, required for HS2.
Geotextiles and reinforced ground can bridge a certain amount, but until a clearer picture is
available of the potential scale of disruption, it is not possible to be confident of their effectiveness.
SRGS refers to this in 8.8.2, as providing ‘some spanning of poor ground’. It is very uncertain that this
would be sufficient in wet-rockhead areas – the examples below show the more radical uses to date.
High speed railway track over karst foundation, LGV Est, Lorraine, France.27 During
construction of this high speed railway line, near La Croix-sur-Meuse in the East of France,
the contractor discovered cavities in the foundation of a long cutting along the proposed
track alignment. The cavities existed in a karst limestone layer that coincided with the base
of the high speed track structure. The width of the joints in these fractures ranged between
25 Case History Long-term time-lapse microgravity and geotechnical monitoring of relict salt mines, Marston,
Cheshire, U. K. Jamie K. Pringle , Peter Styles , Claire P. Howell , Michael W. Branston , Rebecca Furner , and
Sam M. Toon. Geophysics, Vol. 77, No. 6
0.15 m and 0.20 m. The best design option arrived at was to include geosynthetic
reinforcement to span across any potential foundation voids, and thus minimise any
resulting surface differential deformations. From a design perspective, it was assumed that
the maximum possible cavity diameter that could develop was 0.5 m. Further, to maintain
train speed, it was determined that the maximum surface deformation be limited to a
maximum of 1 mm below the ballast level over any 0.5 m diameter void. The total cost was
about €4 billion; 61% public funds.
The Spanish high speed train from Madrid to the French border28. The geological profile
along the length of the track varies considerably, and in some areas the foundations are
prone to collapse due to the presence of karst formations of limestone and gypsum soils. In
the area of Guadalajara, an exhaustive examination of the limestone foundation strata
confirmed the presence of cavities due to the dissolution of the limestone. Depending on
the diameter of the cavities, three different corrective techniques were employed: refill the
cavities with concrete grout; excavation and construction of concrete slabs; and the use of
geosynthetic reinforcement. All corrective techniques used had to ensure that no
discernable surface deformations would occur if cavities later formed beneath the track
support layers. The geosynthetic reinforcement solution was employed in railway cuttings in
this area where the likely cavity diameters were quite small (≤ 0.5 m).
Total investment : 14,267 million euros: Spanish Government; EU Cohesion Fund: 3,358
million euros; Trans-European Transport Network Funds : 81.5 million euros; European
Investment Bank 2,500 million euros
12. Distribution of Risks and Responsibilities
The responsibility for assessing the suitability of a site for any purpose rests with the developer, as
does the responsibility and subsequent liability for its safe development and secure occupancy.
Whilst conditions may be unforeseen the question as to whether they were unforeseeable arises29.
The Latham Report stated that risk …can be managed, minimised, shared, transferred or accepted; it
cannot be ignored.30
When it comes to the risk of unforeseen ground conditions, a huge area of risk, who will be bearing
that risk for HS2? HS2 Ltd, the consultants and engineers, indemnity insurance, CBSCB, Rail Track?
No, the taxpayer.
This first risk assessment for HS2 was drawn up in 200931, and updated in 201132. Although (3.1.8)
Civil engineering works were identified as the driving cost factor in construction averaging, 73% for
the six European lines studied, ground conditions do not appear in the top ten risks. In the HS2 route
wide construction risk register, Geotechnical costs (owing to Uncertain Ground Conditions resulting
in the cost of ground improvements higher than expected, and the cost of alternative design – risk
level 3) had a most likely cost of £100 million in 2009, £200 million in 2011 (maximum £300 million).
Salt subsidence had not been recognised at that time.
28 28 http://www.tencate.com/amer/Images/Reinforced%20Soil%20Case%20Studies_tcm29-19401.pdf
29 Civil Engineering & Mining Related Geohazards – A Clients Guide To The Regulatory Process Leigh Sharpe, The Coal
HS2 Phase Two Risk analysis for the Economic Case Technical documentation does not refer to
geotechnics33. Nor are ground conditions considered in the High Speed Rail: Consultation on the
route from the West Midlands to Manchester, Leeds and beyond. 34
The National Audit Office ‘Progress with preparations for High Speed 2’ (2016) found that the £55.7
billion funding package does not cover funding for all the activity needed to deliver the promised
growth and regeneration benefits which is the responsibility of local authorities. There is risk that
these benefits will not materialise if funding cannot be secured. Any development associated with
HS2 in this area will also need geotechnical investigation, monitoring and remediation. This is
unlikely to be forthcoming.
34 Sustainability Statement Volume 1: main report of the Appraisal of Sustainability A report by Temple-ERM for HS2 Ltd