15.215.415.615.816.016.216.416.616.817.0

766

767

768

769

770

771

772

773

-400 -300 -200 -100 0 100 200 300 400

Tem

pera

ture

(°C

)

Con

duct

ivity

(µ S

/cm

)

Distance (m) along survey area

A: 2015

Conductivity (µ S/cm)Temperature (°C)

16.5

17.0

17.5

18.0

18.5

19.0

792

794

796

798

800

802

804

806

-400 -300 -200 -100 0 100 200 300 400

Tem

pera

ture

(°C

)

Con

duct

ivity

(µ S

/cm

)

Distance (m) along survey area

B: 2016

Conductivity (µ S/cm)Temperature (°C)

Direction of drift

Direction of drift

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

-400 -300 -200 -100 0 100 200 300 400

Dep

th (m

)

Distance (m) along survey area

Sonar Depth 2016Stick Depth 2016Average (offical) DepthSonar Depth 2015Stick Depth 2015

-0.20

-0.15

-0.10

-0.05

0.00

0.05

-400 -300 -200 -100 0 100 200 300 400

Bou

guer

Gra

vity

(mG

als)

Distance (m) along survey area

2012201320152016

Key:

Ongoing subsidence in Marston, Northwich: Using time-lapse microgravity and conductivity-depth-temperature surveys to explore causes and mechanisms.

Keele University, School of Physical Sciences and Geography, William Smith Building, Keele, Staffordshire, UK. E-mail: c.j.rowell@keele.ac.uk and j.k.pringle@keele.ac.uk

BackgroundSubsidence as a result of mining in the UK has occurredthroughout history. It varies in severity, from topographicdepressions to catastrophic surface collapse. Extensivesubsidence occurs due to salt dissolution and mining/brineextraction. Historically, salt (NaCl) has been considered aprecious commodity, with the earliest rock salt explorationin Europe dating back to Neolithic times (Weller &Dumitroaia, 2005).

90% of the UK’s total salt output is within the TriassicMercia Mudstone Group which underlies much of Cheshire.In Northwich, Cheshire and the surrounding areas, the longhistory of salt mining and brine pumping has resulted insignificant surface subsidence, leaving damagedinfrastructure and structurally unsafe domestic housing inits wake. Increased understanding of the mechanisms ofsubsidence, would allow the nature, magnitude, locationand timing of subsidence occurrences to be betteranticipated, and would enable preventative measuresand/or mitigation to be undertaken.

10 years of microgravity surveys, in addition to 20 years oftopographical surveys, provided an insight into areas ofsubsidence which could not have been anticipated fromsite/mine abandonment plans and shaft data alone.

The potential flood risk involved in further subsidence alongthe Marston stretch of the canal, especially in view of theabsence of canal locks for ~40km, are of particularconcern. The presence of voids and zones ofunconsolidated material (≤50m depth) were identified at theLion Salt Works, closed in 1986 (Adams et al.,1992).Topographic monitoring along the canal was regularlyconducted by British Waterways between 1998 and 2008,and remedial action was taken when subsidence wasobserved: ~30cm of concrete and fill material was added tothe banks (Pringle et al., 2012).

By collating existing borehole and access shaft information,a schematic cross section of the survey area wasproduced, and geophysical monitoring of the developmentof subsurface void spaces was conducted by Pringle et al.(2012): microgravity readings were collected between 2002and 2011 [Figs. 2&3] during the summer seasons.

Aims• To collect microgravity data along the survey area and compare it to previous years’ data in order

to ascertain if the trend identified by Pringle (2012) continues.

• To obtain conductivity-depth-temperature (CDT) measurements along the survey area in order to establish any connection beneath the canal and underlying/adjacent mines, given the prior existence of a gravity anomaly along the Canal and the remedial works which have been undertaken in order to maintain the towpath.

Methodology, Results & Discussion

Dep

th

Repeat microgravity surveys [Fig. 8] between 2012 and 2016show that the trend identified by Pringle (2012) continues: Anegative gravity anomaly was observed in the 2012-2016 data[Figs. 5] which was consistent with, and in the same locationas, that observed in the 2002-2012 data [Figs. 5]

This is in support of Pringle’s (2012) finding that maximumsurface subsidence is on mine margins, as opposed to thecentre of abandoned mines.

Further numerical 2D modelling should be applied taking theDepth results (below) into consideration, in order to betterinform geophysical data interpretation of this stretch of thecanal in the future.

Continued ResearchGeophysical:

Monitoring the survey area in Marston, and potentially along other margins around / •between mines where similar subsidence may occur, using time-lapse microgravity Application of GPR to the canal itself (towed by boat) and (should access be permitted) •to boreholes. This will allow an image of the canal base and subsurface to be obtained.

Hydrogeological:Wharmby (1987) makes numerous references to mine owners drilling through toneighboring (disused) mines in order to expand their brine production in the past. Furtherdesk research conducted shows that Neumann’s Flash collapsed repeatedly between1873 and the 1930s through both the top and bottom Northwich halite beds/mines [Fig.11].Given the ambiguity regarding connections between the (now collapsed) mines beneathNeumann’s Flash and surrounding mines (especially the Marston Old Mine), the

AcknowledgementsI would like to thank Jamie Pringle for his supervision, for supplying previous years data and his support andencouragement throughout the project. I would also like to thank Jess Fulton, Nicholas Mason, Peter Rowell,Liz Mighall, David Blackhurst and Oliver & Christopher Bagley for assisting me in the collection of field data,Cheshire Brine Subsidence Compensation Board – especially Pauline Cooke (CBSCB) for her support andinspiration – for project support and ACORN / FRO for funding.

Cathrene J. Rowell & Jamie K. Pringle

• The results of the 2015 depth tests [Fig. 6] were interesting:Contrary to the intended use of the stick as a tool for testingthe reliability of the sonar equipment [Fig. 9b], it served tohighlight the depth of the sediment when compared to thesonar reading. At ~75m, the stick was unable to reach thebase of the canal, and therefore a minimum depth of >2.87mwas recorded.

• In 2016, the methodology was changed in order toaccommodate (longer) stick depths at all survey stations.The results of these depth tests [Fig. 6] highlighted:• Little change in the sediment thickness between 2015 &

2016• A much greater canal depth (4.54m) at the location of the

microgravity anomaly. This is much greater than the 1.68mofficial depth at this point.

Mic

rogr

avity

There is no doubt that subsidence has occurred, and continues along the survey areaand canal in Marston, Northwich:

2012• -2016 time-lapse microgravity results are consistent with those presented byPringle (2012) in terms of both the overall trend and the negative gravity anomaly.

Measurements of the depth to canal base are congruous with the microgravity data, •with a recorded depth of 4.54m corresponding to the largest negative gravity anomaly.

Conductivity and temperature results also present anomalies in the area of concern •identified by the microgravity surveys. These appear to reverse seasonally: more research is required in order to obtain a better understanding of this.

Given the combined microgravity and CDT results, there is undoubtedly a connection between the canal and one (or more) of the neighbouring/underlying mines, however further research, investigation and refinement of methodologies is required in order to establish the source and mechanism of this instance of subsidence and the nature and extent of any hydraulic continuity which may exist.

Conclusions

Con

duct

ivity

& T

empe

ratu

re

Figure 3:Long-term (10-year) time-lapse microgravitymonitoring data of thesouth bank section of theTrent and Mersey Canal.Data have been detrendedto remove longerwavelength variabilities.Adapted from Fig. 8,Pringle et al. (2012)

Figure 11: Marston & surrounding mines and Neumann’s Flashes. Created with informationcompiled from Wharmby (1987), Hewitson (2015) and the Salt Union Abandoned Mine Plan (1925).©Crown Copyright Database 2011

References• ADAMS, S. & HART, P. A. Ground conditions at Lion Salt Works site, Marston, Cheshire. Proceedings of the 4th International

Conference on Ground Movements and Structures, 1992 Cardiff, Wales. Pentech Press, 443-458.• ARUP GEOTECHNICS, 1990. Review of mining instability in Great Britain, Stafford brine pumping. Arup Geotechnics for the

Department of the Environment, 3.• EARP, J. R. & TAYLOR, B. J. 1986. Geology of the country around Chester and Winsford, Natural Environment Research.• HEWITSON, C. 2015. Saltscape Project A8: The salt fields investigating the salt production landscape in Northwich’s historic

saltscape, 2015-001. Cheshire West and Chester.• IGHT, T. S., LICHT, S., BEVILACQUA, A. C. & MORASH, K. R. 2005. The fundamental conductivity and resistivity of water.

Electrochemical and Solid-State Letters, 8, E16-E19.• PRINGLE, J. K., STYLES, P., HOWELL, C. P., BRANSTON, M. W., FURNER, R. & TOON, S. M. 2012. Long-term time-lapse

microgravity and geotechnical monitoring of relict salt mines, Marston, Cheshire, UK. Geophysics, 77, B287-B294.• SALT UNION. 1925. Abandoned Mine Plan, Number 8123, Cheshire.• WELLER, O. & DUMITROAIA, G. 2005. The earliest salt production in the world: an early Neolithic exploitation in Poiana Slatinei-

Lunca, Romania. Antiquity, 79, Online (http://antiquity. ac. uk/projgall/weller/).• WHARMBY, P. 1987. Survey of abandoned salt mine workings and brine shafts in Cheshire. Cheshire County Council Report.

Additional map: SJ 6675 & 6775 Sheet

Figure 1: Location map of the Marston field area (box) north of thetown of Northwich, Cheshire, with U. K. location map (inset). Alsomarked are topographic survey sample positions along the Trentand Mersey Canal, subsidence-prone areas, and the gravity basestation position. Background image provided by OrdnanceSurvey/EDINA service. © Crown Copyright Database 2010. Adaptedfrom Fig. 1, Pringle et al. (2012)

MethodologyThe • Scintrex™ CG-5 automated microgravity meter was implemented for all surveys [Fig. 8].A minimum of three base station [Fig. • 1] readings were taken at the beginning and end of each day to ensure reliability. The microgravity survey took three measurements of • 90 seconds (2013, 2015 & 2016) and 75 seconds (2012) each per station to improve data reliability. Data was collected in 2015 & 2016 by myself, and 2012 & 2013 was provided by Jamie Pringle.Station spacing was initially • 20 m, and was reduced to 5 m over the main area of concern highlighted by Pringle et al., (2012).

Methodology• 3 depth readings were taken, roughly in the centre of the canal and level with each survey station [Fig.

9c], using a hand-held sonar device [Fig. 9b]. A riverboat and a canal boat were used to access thecanal [Fig. 9d/e] .In• 2015, reliability of the depth readings was verified by using a 3m stick with measuring tape attachedto it at around ~10% of survey stations. In 2016 A further measurement was taken using a telescopicpole (sturdier and longer than the previous year (5m)) [Fig. 9a] at all survey stations. Notes were madeon “softness” / presence of silt on the canal floor – pushing it down till it reaches the hard base.

Methodology• 3 readings were taken in the top surface waters of the canal at each station using the WTW MultiLine

P4 [Fig. 10a].Each• reading [Figs. 10b-d] . was taken over a period of 20 seconds, and the equipment was rinsedusing Base Station water (local tap water) and dried with a microfiber cloth between each station.

• ~10% of survey stations revisited and re-surveyed to ensure reliability, and notes were made onproximity of boats, weather changes and the direction of any canal drift on that day.

EC readings along the canal were higher that the conductivity •of typical drinking water (50-500 μS/cm) but much lower than the conductivity sea water, (~5x104 μS/cm) in both years (Light and Licht et al., 2005):

2015 ranged from ̴ 767 μS/cm to ~773 μS/cm, with an average EC of 771 μS/cm overall. 2016 ranged from ~794 μS/cm to ~805 μS/cm, with an average EC of 799 μS/cm overall.

A negative correlation between the temperature and •conductivity in 2015 [Fig. 7a], which shows temperature increasing as the conductivity of the water decreased at ~100m – 150m. A negative correlation between the temperature and •conductivity in 2016 is also apparent [Fig. 7b] , which shows temperature decreasing as the conductivity of the water increased at ~0m – 100m. The variations in the location of these anomalies in relation to •the gravity anomalies [Fig. 5] could be explained by the direction of the canal drift on the survey date which are marked on [Fig. 7a&b] .

Figure 5: Long-term (5-year) time-lapse microgravity monitoring data of the south bank section of the Trent and Mersey Canal. Data have been detrended to remove longer wavelength variabilities.

Figure 6: Results of the depth tests of the centre of the canal along the survey area (2016 & 2015).

Figure 7: Conductivity and temperature variations along the survey area Autumn 2015 (A) & Spring 2016 (B). The direction of the drift of the canal water is market on the graphs.

A

B

Figure 2: Marston site map showing current infrastructure,geophysical sample positions, abandoned salt mines and accessshafts (initials), and borehole positions (see key). Backgroundimage provided by Ordnance Survey/EDINA service. ©CrownCopyright Database 2010. Adapted from Fig. 4, Pringle et al.(2012)

a

b

c

d

e

Figure 9: a: Measuring Rodimplemented in 2016, b: NorcrossH22PX handheld Digital DepthSounder, c: Controlling boat andensuring survey points correctlyaligned, d: River boat used for datacollection 2015, e: Canal boatused for 2016 collection.

Figure 8: a:, Leica Pinpoint R100 and Prism & Pole in use b:golf umbrella implemented to protect the Scintrex™ CG-5gravimeter

a

b

proximity of the mines to eachother, the investigation of apotential hydraulic continuitybetween the relict mines andNeumann’s flash should beconducted:

Water• samples should beobtained from the canal(regularly spaced between ~0-200m, where the negativegravity anomaly appears [Fig. 6])and the geochemistry of thesamples compared with samplesobtained from Neumann’s Flash.

By• closely monitoring rainfalland temperature via a weatherstation erected within one mile ofthe survey area, samples shouldbe collected in different seasons,and under different weatherconditions (e.g. after a period ofheavy rainfall), in order toascertain whether such ahydraulic continuity exists.

Figure 4: Geotechnical 2D models of survey area (bottom) with 2011corrected gravity data and model-calculated gravity (top). This modelshows collapsing mines to best-fit 2011 corrected gravity data. NB:calibrated to boreholes, mineshaft, mine abandonment plans and rockdensities. Adapted from Fig. 9(B) Pringle et al. (2012))

After corrections, analysis and modelling were complete, it was established that the deepeningnegative gravity anomaly observed [Fig. 3] (and associated subsidence) was located at the minemargins, as opposed to above the relict mines. The 2D geotechnical model of survey area [Fig. 4],illustrates how, although the microgravity response generated largely supports the intact minesbeneath the surface, the (best-fit) modelling of several upwardly propagating voids, resulting fromnatural dissolution and collapsing mines, to the 2011 data, provides a plausible explanation for thesubsidence which has occurred along this stretch of the canal.

Considering the infrastructure in and around Northwich, inaddition to the increased development during the pastdecade, it is essential that identification of areas at risk ofsubsidence is made to ensure the stability of current andfuture infrastructure e.g. buildings, railways, canals.

Pringle’s (2012) research shows that time-lapsemicrogravity is an effective tool to investigate and monitorsubsidence resulting from natural dissolution, as well asthe anthropological salt and brine extraction which hastaken place in the town of Northwich, Cheshire.

Figure 10: a:, WTW MultiLine P4,notebook and survey point #6, b - d:Measurements being taken from the canalbank.

ba

c db