You are on page 1of 12

2012-013

2013
research-articleResearch ArticleXXX10.1144/qjegh2012-013Deoxygenated Gas in the Lambeth GroupT. G. Newman etal.

Deoxygenated gas occurrences in the Lambeth Group of


central London, UK
T. G. Newman1,2*, R. C. Ghail1 & J. A. Skipper3,4
1Department of Civil and Environmental Engineering, Imperial College London, London SW7 2AZ, UK
2Thames Water Utilities, London Tideway Tunnels, The Point, 7th Floor, 37 North Wharf Road, Paddington,
London W2 IAF, UK
3Geotechnical Consulting Group, 52A Cromwell Road, London SW7 5BE, UK
4Department of Mineralogy, Natural History Museum, Cromwell Road, London SW7 5BD, UK

*Corresponding author (e-mail: t.newman10@imperial.ac.uk)

Abstract: Deoxygenated air poses a serious, life-threatening hazard (confined space hypoxia) for engi-
neering projects in London, particularly within the Upnor Formation of the Lambeth Group. This paper
reviews its causes and postulates that it was induced by regional-scale dewatering and drawdown of
the Lower Aquifer during the industrial growth of London, bringing air into contact with these sediments,
which became oxidized. Ensuing post-industrial recharge and resaturation resulted in accumulations of
often compressed deoxygenated air, trapped beneath overlying impermeable clay strata. Historically,
glauconite has been considered to be the mineral most likely to remove oxygen from the trapped air,
but it lacks potency as a reducing agent and remains unaltered in many oxidized sediments. Pyrite and
organic carbon are both plausible, but rare. Green rust, a mixed Fe(II) and Fe(III) layered double hydrox-
ide, is more likely, as it rapidly oxidizes on contact with air and is only briefly observed in fresh core
samples. Two key ground engineering hazards are, therefore, identified: the misidentification of risk by
reliance on observations of glauconite; and the likelihood of encountering pressurized deoxygenated air
within Lambeth Group sediments in underground projects following short- or long-term changes in the
groundwater level.

Confined space hypoxia (Zugibe etal. 1987), caused by deoxygen- 1970); however, the cause of the reduction in oxygen was never
ated gas, is well known within the water and sewerage industry and identified.
past fatalities have led to increasingly stringent Health and Safety In 1984 an oxygen-deficient atmosphere was found at the start of
legislation for working in confined conditions underground (HSE each day shift (E. Woods & A. Miller, pers. comm.) during construction
2009). Nevertheless, even with good compliance, hypoxic ground of the Effra Storm Relief Sewer, south London. The 183m internal
conditions can occur and rapidly lead to unconsciousness and even diameter tunnel was driven beneath Brockwell Park, Dulwich [National
death. It is, therefore, imperative to understand and identify the Grid Reference TQ 532 174], as a day-works only contract, with the
conditions under which confined space hypoxia occurs so that the ventilation switched off at the end of each day shift. Low oxygen levels
risk can be correctly quantified. were encountered at the start of each day shift whilst tunnelling through
This paper highlights a specific type of confined space hypoxia the Streatham Fault Zone (Fig. 3), which juxtaposes the London Clay
that occurs within Lambeth Group sediments in central London Formation and Lambeth Group sediments, particularly the Upnor
(Figs 1 and 2), in which air with very low oxygen content is encoun- Formation. Normal levels of oxygen were restored only after a period
tered underground, often under an excess pressure and occasionally of forced ventilation, for which the capacity was increased for the
with elevated levels of carbon dioxide. remainder of the tunnel drive.
The most serious incident to date took place in August 1988,
when two fatalities occurred shortly after construction of a Thames
Water Ring Main (TWRM) shaft at Stoke Newington [TQ 532
Case histories of deoxygenated gas 187], north London. The shaft was 465m deep with an 119m
during construction internal diameter and had been completed 26months before the
incident. An engineer and assistant were both asphyxiated within
Normal dry air consists of 780% nitrogen, 209% oxygen and 09% the confined space of a sealed section of pipework extending from
argon by volume, with other gases, principally carbon dioxide, mak- a high-integrity valve into a closed tunnel heading within the Upnor
ing up the remaining 01% (Doyle 2001). Oxygen deficiency is a Formation. The subsequent Crown Court trial (CNPlus 1991) con-
dangerous condition that arises when human tissue becomes cluded that glauconite, within the groundmass, was the probable
deprived of oxygen, often with fatal consequences. In the UK, an cause of the hypoxia.
underground excavation is considered oxygen-deficient when the Within the same TWRM contract, poor air quality was subse-
concentration of oxygen falls below 190% (BSI 2011). quently encountered during the tunnel drive from Coppermills [TQ
One of the earliest documented incidents of deoxygenated gas 535 188], Waltham Forest, north London, to this shaft. Correspondence
during tunnelling operations in London occurred during construc- from the Resident Engineer (P. Hemmings, pers. comm.) within the
tion of the London Underground Limited (LUL) Victoria Line project team noted that air in this section, through the Upnor
through Lambeth Group deposits (Morgan & Bartlett 1969; Clark Formation, contained as little as 11% oxygen. A number of

Quarterly Journal of Engineering Geology and Hydrogeology, Vol. 46, 2013, pp. 167177 2013 The Geological Society of London
doi: 10.1144/qjegh2012-013
Published Online First on March 25, 2013
168 T. G. Newman et al.

FORMATION Member / Beds Description Appearance

LONDON CLAY N/A Silty to slightly sandy CLAY

Swanscombe Sandy shelly CLAY


HARWICH
Oldhaven Clayey GRAVEL / SAND

WOOLWICH Upper Shelly Beds Shelly CLAY

Blue grey mottled sandy


READING Upper Mottled Beds CLAY AQUICLUDE

Laminated Beds Laminated fine sandy CLAY


WOOLWICH
Lower Shelly Beds Shelly CLAY MID-
LAMBETH
Blue orange red mottled HIATUS
READING Lower Mottled Beds CLAY with sand and gravel
AQUICLUDE

Nitrogen gas
UPNOR N/A Clayey SAND with gravel
under pressure
Silty fine SAND clayey
N/A towards base
THANET SAND
Bullhead Bed Clayey, sandy flint GRAVEL
HYDRAULIC
CONTINUITY
Haven Brow
of Lower
SEAFORD Mod weak to weak white Aquifer
Cuckmere
CHALK CHALK

Belle Tout

Fig. 1. The geological succession for London.


5

4
M11
17
17

M1 28
8
OXYGEN DEPLETION DURING
SHAFT CONSTRUCTION
2 M25
TWRM 4
190000
OXYGEN DEPLETION DURING
TUNNEL CONSTRUCTION
29
29
M25 COPPERMILLS
AUDIBLE GAS MOVEMENT DURING 1 STOKE NEWINGTON
TUNNEL CONSTRUCTION CTRL X-P
16
16
1a
a
M40 1 NLFRS
GASSING BOREHOLE DURING TWRM EXT.
GROUND INVESTIGATION
ABBEY MILLS
NO OXYGEN DEPLETION DURING NEW RIVER HEAD
SHAFT CONSTRUCTION
BH IOD30R
CTRL INCIDENT BH 1066D
180
30 000

31
4 2
3 1 BH 2071
5 M4
4a
4a

14
14
HONOR
OAK
M25
BRIXTON

EFFRA
1
SRS

M25
12
1 2
3
2
LONDON

M20
11
11

510000 520000 530000 540000 2013 Google

Fig. 2. Location of ground gas occurrences during tunnelling and borehole investigations. Grid lines have a separation of 10km and are oriented north
south (up and down the page) and eastwest (across the page). Map data @ Google 2013.
DEOXYGENATED GAS IN THE LAMBETH GROUP 169

6+$)7

6+$)7
6+$)7'
7:50(;7
(16,21

%52&.:(//3$5.

675($7+$0)$8/7=21( 6+$)7&



(/(9$7,21 P2'
6+$)7'
:DWHU :DWHU 6+$)7&
6+$)7 6+$)7 LQIORZ VHHSDJH
OPLQ


2[\JHQGHSOHWHG DLU
&RPSUHVVHGDLU SVL


7XQQHO(QJLQHHULQJ*HRORJLFDO'HVFULSWLRQ ,QWHUSUHWHG*HRORJ\
6WLII/RQGRQ %OXH &OD\ /RQGRQ&OD\)RUPDWLRQ

'DUNJUH\FOD\ZLWKVKHOOVDQGOLPHVWRQHEDQG 8SSHU6KHOO\%HGV

'DUNJUH\VLOW\VDQG\FOD\ZLWKVKHOODQGORDPEDQGV /DPLQDWHG%HGV

'DUNJUH\WREODFNVKHOO\FOD\ZLWKFHPHQWHGVKHOOV /RZHU6KHOO\%HGV

6WLIIPRWWOHGEOXHEURZQUHGDQG\HOORZFOD\ /RZHU0RWWOHG%HGV

3HEEOHVLQEURZQUHGVDQG\FOD\PDWUL[ /RZHU0RWWOHG%HGV SHGRJHQHWLFDOO\DOWHUHG8SQRU)RUPDWLRQ 

)OLQWSHEEOHVLQGDUNJUHHQFOD\H\WRVDQG\VLOWPDWUL[ 8SQRU)RUPDWLRQ

/LJKWJUH\ILQHVDQG 7KDQHW6DQG)RUPDWLRQ

Fig. 3. Schematic geological section along the Effra Storm Relief Sewer tunnel.

construction engineers (E. Woods, G. Howells & A. Miller, pers. 98% nitrogen, less than 12% oxygen, and elevated levels of carbon
comms.) working on this section gave accounts of gas escaping dioxide, in excess of 10%. Rates of gas emission from each borehole
through the final invert of the shaft at Coppermills, seen as bubbles were between 16 and 50m3s1 indicating a pressurized gas source.
rising through water ponded on the shaft floor during excavation. Monitoring of the boreholes continued for a year, over which time
The oxygen-depleted gas was encountered in the section of tunnel the levels of oxygen were recorded consistently below 10%, with
in which the invert and base slab were constructed within the Upnor carbon dioxide levels between 10 and 20% (Fig. 6). Lewis &
Formation (Fig. 4), a section recorded as dry in piezometer data Harris (1998) proposed a direct link between these gas levels and
collected. The oxygen deficiency was mitigated through ventilation barometric pressure; however, based on results of borehole moni-
and air quality was monitored to ensure safe working conditions. toring during ground investigation for both the north and south
A near miss incident occurred during construction of a cross- TWRM and current Thames Tideway Tunnel scheme, there appears
passage (C. Warren & E. Woods, pers. comm.) between two com- no evidence to support such a link in these data.
pleted running tunnels of the Channel Tunnel Rail Link (CTRL) Almost the entire 282m internal diameter of the 70km long
beneath Hackney Wick [TQ 536 185] in north London. Another NLFRS tunnel was driven through the Upnor Formation. Given the
incident took place during caisson sinking of the TWRM shaft at results of the borehole investigations, clearly defined underground
Honor Oak [TQ 535 175], south London (Newman & Wong 2011). entry procedures and an increased capacity ventilation system were
In both these incidents, an oxygen-deficient atmosphere was enforced to mitigate any risk of injury caused by poor air quality
recorded during construction in the Upnor Formation after a period during construction. However, gas was frequently heard passing
of dewatering and drawdown from the underlying chalk. through the tunnel lining throughout this time (Lewis & Harris
Significant oxygen depletion was also recorded within a num- 1998; G. Howells & A. Sefton, pers. comm.). It is possible that this
ber of borehole monitoring installations during separate ground was in response to the high forward ground pressures induced dur-
investigations for two large tunnelling schemes in north London: ing operation of the tunnel boring machine (TBM).
Thames Waters North London Flood Relief Sewer (NLFRS), The 285m internal diameter TWRM Stoke Newington to New
between Stratford and Stoke Newington; and the TWRM northern River Head (NRH) tunnel, in northcentral London, was also con-
extension tunnel, between Stoke Newington and Finsbury. structed predominantly within the Upnor Formation, which was mostly
During borehole investigations for the section of the NLFRS above the water table of the Lower Aquifer (Fig. 7; Newman etal.
between Victoria Park and Stoke Newington (Fig. 5), samples of 2010). During ground investigation for the tunnel, measurements of gas
gas were taken from borehole installations within the Upnor from within a number of borehole monitoring installations in the Upnor
Formation. Laboratory analyses found that the gas contained up to Formation indicated up to 24% carbon dioxide and less than 1.0%
170 T. G. Newman et al.

NW SE
40 40
STOKE NEWINGTON
SHAFT
30 30

20 20

COPPERMILLS
SHAFT
10 10
ELEVATION (m OD)

ELEVATION (m OD)
0 0

GASSY GROUND
-10 DRY TUNNEL
-10
DRY DRY
DRY DRY TUNNEL
GASSY GROUND

-20 -20

-30 -30

2 km

KEY TO GEOLOGY AND SYMBOLS

MADE GROUND & ALLUVIUM RIVER TERRACE DEPOSITS LONDON CLAY FORMATION LAMBETH GROUP (EXCL. UPNOR FORMATION)

3('2*(1,=(' 83125 UPNOR FORMATION THANET SAND FORMATION WHITE CHALK SUBGROUP

POSITION OF BOREHOLE MONITORING INSTALLATION AND WATER LEVEL RISE INTERPRETED PIEZOMETRIC LEVEL

Fig. 4. Schematic geological section between Coppermills and Stoke Newington.

oxygen. Although nitrogen was not measured, monitoring of the bore- and in Tower Hamlets [TQ 535 180] in 2009. In these cases the
holes was undertaken over a 12month period. The oxygen levels (Fig. deoxygenated gas contained elevated levels of carbon dioxide.
8) were more variable than those in the NLFRS data, often with a sharp At Romford [TQ 551 181], pockets of air containing severely
decrease from a concentration between 120 and 150% to one below depleted oxygen, at 01%, and raised carbon dioxide levels as high as
50%, remaining stable at that level thereafter and displaying no clear 26%, were recorded during subsequent monitoring of standpipe instal-
link between the gas levels and barometric pressure. As in the NLFRS lations. At Tower Hamlets, 87% oxygen and 06% carbon dioxide
tunnel, gas was heard passing through the tunnel lining during construc- were recorded. At both locations drillers recorded blowing sand dur-
tion (V. Yuan, pers. comm.). ing borehole excavation and distinct voids were observed in rotary core
A similar incident was experienced during ground investigation samples recovered. These are presumed to represent gas migration
for the Crossrail scheme in Londons docklands area, whereby gas pathways; material surrounding the voids was noted to be patchily oxi-
was encountered escaping from Borehole IOD30R (location confi- dized, although the origin of the gas is unclear at this stage.
dential; refer to Fig. 2 for location) upon its intersection with the
Upnor Formation (U. Lawrence, pers. comm.). Subsequent gas
monitoring of this indicated significantly reduced levels of oxygen
and slightly elevated carbon dioxide, but no obvious associated
Causal mechanisms
effects from changes in barometric pressure. Physical processes; groundwater
At each of the above-mentioned locations, deoxygenated gas
levels and lithological controls
was encountered within the Upnor Formation above the water table
of the Lower Aquifer (Thanet Sand Formation and Chalk Group). The Upnor Formation is usually hydraulically connected with
However, deoxygenated gas has also been recorded in sand beds the Lower Aquifer beneath London (Simpson etal. 1989). Prior
within the Laminated Beds of the Woolwich Formation, in bore- to significant groundwater pumping, the Lower Aquifer was
holes for proposed residential developments in Romford and Little under artesian pressure (Fig. 9), confined throughout much of
Ilford, Essex, in 2005 and 2008 respectively (P. Gawne, pers. London by the overlying Lower Mottled Beds. As discussed by
comm.). This has also been recorded in the current Thames Aldiss (2012), these are the Lower Mottled Clays of Ellison
Tideway Tunnel investigation boreholes along a section of the (1983) and the pedogenically altered, mottled, Upnor
Victoria Embankment, central London, at Temple [TQ 531 180] Formation. However, for the purposes of this paper, the term is
DEOXYGENATED GAS IN THE LAMBETH GROUP 171

NW SE
30
278D/2

SANDFORD

278D/7
TERRACE

278D/9
20 HACKNEY

278C/15
DOWNS VICTORIA
PARK

278C/32
10 ABBEY 10
LEE VALLEY MILLS
ELEVATION (mOD)

0 0
DRY

DRY DRY

-10 DRY DRY -10


DRY DRY DRY
DRY
DRY

-20 -20
?

TUNNEL
-30 -30

1.5 km -40

KEY TO GEOLOGY

MADE GROUND / ALLUVIUM RIVER TERRACE GRAVEL LONDON CLAY FORMATION LAMBETH GROUP (EXCL. UPNOR) UPNOR FORMATION

3('2*(1,=(' 83125 )250$7,21 GRAVELLY UPNOR FORMATION THANET SAND FORMATION WHITE CHALK SUBGROUP

GEOLOGICAL FAULT (WITH THROW) POSITION OF BOREHOLE INSTALLATION WITH WATER LEVEL RISE PIEZOMETRIC LEVEL BOREHOLE

Fig. 5. Schematic geological section between Stoke Newington and Abbey Mills (refer to Fig. 4 for key to geology).

1035
Oxygen
100
Carbon Dioxide
Nitrogen 1030

Barometric Pressure
Concentration / Vol. %

1025
10
Pressure / mbar

1020

1
1015

1010
0.1
29-May-1995

21-Aug-1995

18-Sep-1995

13-Nov-1995

11-Dec-1995
26-Jun-1995

16-Oct-1995
1-May-1995

24-Jul-1995

5-Feb-1996

4-Mar-1996
8-Jan-1996

1-Apr-1996

Date

Fig. 6. Monitored gas levels in North London Flood Relief Sewer borehole 278D.

derived from the system used in civil engineering site investiga- cone of depression beneath west central London (Marsh & Davies
tions, primarily in London and based on Page & Skipper (2000). 1983). From 1940, the rate of abstraction from the Lower Aquifer
Over-abstraction of the Lower Aquifer, primarily from the reduced significantly with the destruction of a large number of
Chalk, began in the early 1800s at rates rising steadily from approx- abstraction wells during World War II and the control of abstraction
imately 9106m3a1 to a peak of 83107m3a1 by 1940 (Hurst & by the 1945 Water Act (Wilkinson 1985). Escalating costs, owing to
Wilkinson 1986). This led to a significant lowering of the ground- the increasing depth of abstraction required, and a reduction in
water table within the Lower Aquifer and the formation of a large industrial requirement, limited the number of new wells constructed.
172 T. G. Newman et al.

+,*+%85<+,//
 
1(:5,9(5+($'
6+$)7 7+($1*(/ ,6/,1*721 672.(1(:,1*721
 6+$)7 

 
+LJKEXU\+LOO)=
 
1HZ5LYHU+HDG
(/(9$7,21 P2' 

 )= 

 

 

 

 

 

 

P P

Fig. 7. Schematic geological section between Stoke Newington and TW New River Head (from Newman etal. 2010) (refer to Fig. 4 for key to geology).

 





&RQFHQWUDWLRQ9RO



3UHVVXUHPEDU






 2[\JHQ

$WPRVSKHULFSUHVVXUH
 
2FW

2FW

1RY

'HF

-DQ

)HE

0DU

$SU

0D\

-XQ

-XO

$XJ

6HS

6HS

2FW

1RY

'HF

-DQ

'DWH

Fig. 8. Monitored gas levels in TW New River Head borehole 773A.

None the less, by 1965 continued pumping had lowered the ground- Hammersmith in west London and the section of the north bank of the
water table by more than 70m in the centre of a cone of depression River Thames between Chelsea and Victoria Embankments.
(lower surface in Fig. 9) (Simpson etal. 1989). This was the lowest By the late 1980s, the continued rise in groundwater level
level recorded and equated to 103109m3 of water abstracted from was recognized as a risk to deep tunnels, deep foundations and base-
more than 121010m3 of the Lower Aquifer (Flavin & Joseph ments by changing the engineering properties of the ground and poten-
1983) over an area greater than 200km2 (Marsh & Davies 1983). tially causing flooding. In response, the GARDIT (General Aquifer
The centre of the cone of depression was located close to Charing Research Development and Investigation Team) strategy was launched
Cross [TQ 530 180] and Trafalgar Square [TQ 530 180] but a sig- in 1998 to control the rising groundwater by abstraction of 70Ml per
nificant area of drawdown occurred within the Lee Valley, near day (256104m3a1) on a schedule up to 2004. However, Thames
Thames Waters Abbey Mills Pumping Station site [TQ 539 183]. Water Plc and Environment Agency groundwater modelling demon-
Since 1965, the rate of groundwater abstraction from the Lower strated in 2000 that this rate was too high (Jones 2007) and should be
Aquifer decreased by between 30 and 50%, from 62107m3a1 in reduced to 50Ml per day (183107m3a1). Continued monitoring
1965 to 43107m3a1 by 1982 (Wilkinson 1985). This was allied with demonstrates the success of GARDIT and the groundwater levels are
the demise of heavy industry, the largest consumer of groundwater, and now broadly stable throughout central London (Jones 2007).
resulted in a rise in the Lower Aquifer groundwater table. Estimates of The drawdown of the Lower Aquifer is believed to have caused
the rate of rise vary, from 15ma1 (Simpson etal. 1989) to 25ma large volumes of permeable, saturated Lambeth Group sediments to
1(Lucas & Robinson 1995), and a more recent estimate of 30ma1 drain, and air to enter the sediments through windows in the over-
(Jones 2007). The greatest recovery is within the central London cone lying London Clay or via large borehole adits, leading in places to
of depression, which rose 20m by 1988. By 1994, the cone of depres- sediment oxidation and the production of deoxygenated air (Flavin
sion had also significantly reduced in extent, becoming confined to an & Joseph 1983). The GARDIT-controlled incomplete recharge of
area between Hammersmith and central London, mainly north of the the Lower Aquifer has trapped and pressurized this deoxygenated
River Thames. Most of the change between the lowest recorded levels air beneath clay beds of very low permeability within the Lambeth
and the highest rate of recovery encompasses an area between Group sediments, in particular the Lower Mottled Beds.
DEOXYGENATED GAS IN THE LAMBETH GROUP 173

HOUSES OF PARLIAMENT

LIMEHOUSE CANAL
VAUXHALL BRIDGE
W E

BATTERSEA PARK

LAMBETH BRIDGE

LONDON BRIDGE
ALBERT BRIDGE

ABBEY MILLS PS
TOWER BRIDGE
HAMMERSMITH

WANDSWORTH
PUTNEY

48 x VERTICAL EXAGGERATION
-10
ELEVATION (m OD)

-20
-30
-40
-50
-60
-70
-80
-90

22 km

LEGEND PRE-ABSTRACTION 1900 1965 1994 THAMES TUNNEL MONITORING (2010)

Fig. 9. Piezometric levels across London 18001994.

The Upnor Formation and adjacent strata are geographically varia- Glauconite is a compositionally variable iron potassium phyllo-
ble but have a broadly consistent stratigraphic sequence (Fig. 1). The silicate within the mica group and the name is used for all green
base of the Upnor Formation is frequently demarcated by a rounded clay pellets with a glauconitic component (Huggett & Knox
fine to coarse flint gravel, within a typically green, clayey, fine to 2006). It is an authigenic mineral that is associated with faecal
medium sand, with variable quantities of the green mineral glauconite pellets in low depositional rate environments, typically in shallow
(Page & Skipper 2000). The main part of the Upnor Formation often water continental shelf (Odin & Matter 1981). It is generally
consists of a greygreen, silty, fine to medium sand to sandy clay, with considered to form as a result of the partial reduction of iron
some fine to medium (occasionally coarse) black flint gravel, com- (Mackenzie 2005). Under the binocular microscope, glauconite
monly containing variable amounts of glauconite. In some locations, varies from pale bluish green to greenish yellow to dark green
particularly to the north and east of London and as encountered during when fresh. However, when oxidized, it breaks down into iron
the NLFRS and TWRM northern extension, distinct thin to medium oxides, which are typically brownish yellow or reddish brown
interlaminated brown clay and fine sand dominates the unit. (Pettijohn etal. 1987).
The resulting emergence of terrestrial conditions over the area of During recent borehole investigations for Thames Waters Tideway
what is now SE England resulted in the formation of the Lower Thames Tunnel scheme, two boreholes, SR2071, in the River Thames,
Mottled Beds (Page & Skipper 2000; Aldiss 2012). These deposits adjacent to Cadogan Pier, Chelsea Embankment [TQ 527 177] and
are characteristically pedogenically altered, typified by the vivid red, SA1066D, Whitehall Gardens, on the Victoria Embankment [TQ 530
orange and purple hues that are associated with oxidation and other 180], intersected pressurized ground gas (Newman & Hadlow 2011).
pedogenetic processes within soil horizons (Ellison 1983; Page & Both boreholes were constructed using cable percussion techniques and
Skipper 2000). Inspection of numerous borehole core samples and in each case pressurized ground gas was encountered at the intersection
underground excavations across the London area demonstrates that with the gravelly Lower Mottled Beds (pedogenically altered Upnor
these pedogenetic processes have affected the top of the Upnor Formation), at 509m below river bed level and 560m below ground
Formation sediments, in particular the gravel that frequently defines level respectively. In the case of borehole SR2071, gravel was ejected to
the top of this unit (Ellison etal. 1994; Ellison etal. 2004). This zone surface, whereas from SA1066D, gas was recorded venting from the
is distinct from the unaltered material below, as it contains red, borehole for 8months until sealed with a permanent cap. A number of
brown, grey and white, subrounded, fine to coarse flint gravel set return visits were made to this borehole to collect gas samples for labora-
within a light bluegrey mottled orangebrown clay matrix. The tory analysis; between 21 August and 15 October 2009 the recorded
effects of pedogenesis diminish with depth, leading to a gradational oxygen content was never more than 34% (by volume) and usually less
contact with the unaltered material below; where permeable, the than 2%, with no apparent relationship with barometric pressure. Carbon
pedogenic zone is in hydraulic continuity with the Lower Aquifer. dioxide was recorded at 600900ppmv, which is two to three times the
normal atmospheric concentration but not considered unusual.
Chemical processes Flow rate measurements (Fig. 10) were not possible until
2weeks after gas was first encountered, so the initial rate is uncer-
There is a need to identify the chemical processes that lead to the tain. Assuming an exponential gas release model (Table 1) the
reduction in oxygen. Within the construction industry, glauconite initial rate is estimated to have been 120lh1, rapidly declining
is regarded as the reducing agent responsible for loss of oxygen over a 2week period to a background level determined from long-
in confined spaces within the Lambeth Group (CNPlus 1991). term monitoring of between 7 and 22lh1. This implies that up to
174 T. G. Newman et al.



 5HFRUGHG)ORZ5DWH'DWD

*DV)ORZ5DWH OLWUHVSHUKRXU 
&DOFXODWHGIURP*DV3UHVVXUH'DWD
0RGHOOHG*DV)ORZ







          
'D\VDIWHU*DV)LUVW(QFRXQWHU $XJ 

Fig. 10. Gas production from borehole SA1066D.

Table 1. Gas release model for SA1066D

Variable Value

Initial rate of gas release 120lh1


Background rate of gas release 14373lh1
Total volume of gas released from disturbed 13915m3
zone
Thickness of soil layer containing gas 2540m
Porosity of soil layer 3550%
Radius of disturbed zone 1803m
Soil permeability with respect to gas flow (2315)10-6ms1
Gas composition 97605% N2, 2407%
O2, 730160ppm CO2

15500l of gas was initially released from a disturbed zone of c. 2m


radius around the borehole and that up to 500000l was released in
total over 2years. The soil matrix gas permeability, c. 106ms1,
inferred from the background rate, indicates that a fine silt and clay
matrix controls gas migration through this layer.
Both boreholes SR2071 and SA1066D lie within the 1964 cen-
tral cone of depression in the Lower Aquifer. It is clear that a large
volume of Upnor Formation became partially saturated as a result
of this cone of depression. However, the subsequent water table
recovery was also greatest here and it is proposed here that the rise
in the groundwater table since 1965 has trapped and compressed
nitrogen-rich, oxygen-depleted gas against the base of the overly-
ing very low-permeability clays of the Lower Mottled Beds.
Up to 14km of drilling samples have been recovered to date during
the ground investigation for Tideway Thames Tunnel. In a number of
instances, examples of apparently fresh (unoxidized) glauconite have
been observed within pedogenetically altered sediments belonging to
the Lower Mottled Beds (Fig. 11), which have been subject to penecon- Fig. 11. Sediments of the Lower Mottled Beds illustrating pedogenic
temporaneous pedogenic alteration, and/or within core samples of the alteration and unoxidized glauconite.
Upnor Formation left exposed to the air over a number of months. This
implies that the oxidation of glauconite is not a rapid process and that
its decomposition above the water table might not make it the primary Beds within the Reading Formation, usually as disseminated fine sand
mineral that reduced the O2 content of the trapped air. grade material or rarely as gravel-sized lumps. Its oxidation leads to the
Alternative reducing agents are organic carbon and pyrite. Organic formation of sulphuric acid, which reacts with carbonate minerals and
carbon in the form of driftwood fragments is present in some parts of buffers the pH of the groundwater, releasing CO2. The oxidation of both
the Group, often as dispersed small particles but sometimes in large organic carbon and pyrite should, therefore, be associated with elevated
concentrations. Oxidation of organic carbon primarily results in the pro- CO2 levels if carbonate minerals are present. Elevated CO2 levels are
duction of CO2. Pyrite is common in the Laminated Beds of the sometimes associated with incidents of confined space hypoxia, such as
Woolwich Formation and occurs in sand channels of the Upper Mottled at Romford, where organic carbon was also noted to be present, but
DEOXYGENATED GAS IN THE LAMBETH GROUP 175

Table 2. X-ray diffraction analysis of Lambeth Group sediments

Component (wt%) Glauconite* Glauconite Kaolinite Smectite Quartz Goethite Pyrite Carbonate

Lower Mottled Beds 0930 125125 7155 356262 273256 4039 0206 0307
LMB Gravels 159174 197178 1422 213167 473258 3246 0105 0712
Upnor Formation 9165 15984 1218 12585 610148 0310 0104 1118
Thanet Sands 2323 5453 0510 2853 816125 0000 0000 0312

*Point-count data (vol%).


XRD; includes illite and mica.
Includes gypsum.
is one standard deviation.

Green rust is suspected to be a very common mineral, widely


distributed in soils and sediments at a depth of a few metres within
certain marine environments and in waterlogged anoxic gley soils
(Genin etal. 2005). Dissimilatory iron (III)-reducing bacteria
(DIRB), in particular Shewanella putrefaciens (OLoughlin etal.
2007), are metabolically most active in these environments (Parmar
etal. 2001) and form green rust through the bioreduction of Fe(III)
oxides and oxyhydroxides.
Green rust crystals are very small, typically about 10nm across
and 3nm thick, and so have a relatively high surface area, rendering
them particularly susceptible to oxidation. The Fe(II) ions oxidize
to Fe(III) ions, transforming the green rust to iron oxyhydroxides
and iron oxides; that is, the commonly observed brown rust, such as
goethite, lepidocrocite and magnetite (Trolard et al. 2007; OLoughlin
etal. 2007; Nagata etal. 2009). Ferrihydrite usually forms as a dark
brown to yellowish brown intermediate product and may be
observed when oxidation is so rapid that alteration to goethite and
lepidocrocite does not occur (Refait etal. 2003).
Fig. 12. Green rust (bluish surface colour) in fresh core sample from the Simple calculations demonstrate the potency of green rust with
Upnor Formation (photograph by B. Patel). respect to glauconite and other possible reducing agents, as sum-
marized in Table 3. Oxidation of 1kg of glauconite (Bentor &
Kastner 1965) consumes 6.5104kg O2. Assuming a gas pressure
usually the hypoxic gas is normal air with reduced oxygen. However, of 500kPa (slightly above hydrostatic conditions at 40m depth),
chemical tests in samples of Upnor Formation (Table 2), in which many 35% average porosity (this study), dry conditions and normal air
of the incidents occurred, show that pyrite is uncommon and organic composition (21% O2), each cubic metre of Upnor Formation sedi-
carbon is present at very low levels (<1%), typically as isolated wood ment contains up to 235kg free (gaseous) O2 before reduction. The
fragments. These levels are unlikely to be high enough to cause any total mass of material in 1m3 of Upnor Formation is 1690kg,
significant oxygen depletion of air. assuming a particle density of 2600kgm3 and 35% porosity. Each
Fresh core samples from the unaltered Upnor Formation and cubic metre, therefore, requires up to 3615kg (i.e. 213.9wt%), of
Lower Mottled Beds are frequently observed as bluish green (Fig. 12) glauconite to fully reduce this oxygen, demonstrating that it is not
that on exposure to air rapidly changes over a few minutes to hours possible for glauconite to be solely or even principally responsible
to dull greyish green and then, over a few hours or days, to yellow- for the loss of oxygen in underground confined air spaces.
ish orange. These changes are characteristic of the oxidation of Organic carbon and pyrite are both very powerful reducing
green rust minerals (Taylor & McKenzie 1980; Christiansen etal. agents, respectively requiring just 088kg and 44kg per cubic
2009). The reaction is far more rapid than glauconite oxidation and metre of ground. Within the borehole cores sampled, both organic
one that consumes much more oxygen, weight for weight. carbon and pyrite are present but as widely dispersed particles.
Green rust is a family of mixed Fe(II) and Fe(III) layered double Their potency for oxygen reduction is thus constrained by low soil
hydroxides with a pyroaurite-like structure consisting of alternat- permeability and by the very low reactive area arising from a large
ing positively charged Fe(II)/Fe(III), brucite-like, hydroxide layers particle volume to surface area ratio. Their potency is thus much
and hydrated interlayers composed of anions and water molecules reduced by permeability and surface area to volume constraints. In
(OLoughlin etal. 2007; Christiansen etal. 2009; Nagata etal. addition, there is no evidence from incidences within the Upnor
2009). Although green rust has a very wide compositional range, in Formation of elevated CO2 or acidified groundwater that are
a typical ground profile the average composition is 27% Fe(OH)2, expected by-products from the oxidation of these materials.
62% Fe(OH)3 and 11% Mg(OH) (Feder etal. 2005). It is character- Deoxygenation of the air to the observed levels of oxygen from
istically bluish green (Legrand etal. 2004; Trolard et al. 2007), form- typical Upnor Formation would require 1053kg of green rust per
ing in environments close to the Fe(II)Fe(III) transition zone, in cubic metre of ground. Green rust forms as a coating on silt- and
conditions with little oxygen. It forms as free colloidal-size parti- sand-sized quartz grains (much like the normal rust that turns sand
cles within natural groundwater, under weakly acidic to alkaline grains red). Assuming spherical quartz grains with an average
conditions, and has an approximate Fe(II) to Fe(III) ratio of 2:1 diameter of 01mm and a green rust density of 3670kgm3, 62wt%
(Zegeye etal. 2005). Although chemically distinct, it plays a simi- green rust would form a 12m thick coating on each quartz grain,
lar role to that of ferric brown rust in forming a stain or coating on giving green rust a remarkably large, reactive, surface area. These
mineral grains in the sediment. factors make green rust a strong candidate for primary chemical
176 T. G. Newman et al.

Table 3. Potency of reducing agents in the Upnor Formation

Compound Reducing potential Mass fraction (wt%) Grain size (mm) Reactive area* (m2) Required mass Reaction products
(kg kg-1) fraction (wt%)

Glauconite 000065 15984 05 1220 2139 Goethite


Pyrite 080 0104 01 20 02 SO3/H2SO4 aq.
Organic C 267 <<01 3 <3 01 CO2/H2CO3 aq.
Green rust 0022 Unknown 01 23340 62 Goethite

*Grain area perm3 of soil.


Required mass fraction to reduce 235kg O2 per m3 of soil.
Goethite and other iron oxides or hydroxides.
Average size of quartz grains coated by green rust.

cause of deoxygenated ground gas in Upnor Formation sediments. against oxidation (Hansen 1989) without unduly affecting the analyses.
However, it is very difficult to identify in the field as it readily Normal powder X-ray diffraction techniques dry the test specimen
oxidizes and does not leave any distinctive by-products, usually before testing. XRD-PSD (position-sensitive detector) is capable of
only oxidation products such as goethite, which may occur as oxi- real-time analysis and is being used alongside infrared spectroscopy
dation of other iron-rich minerals such as pyrite and glauconite. to characterize the reactants and reaction rates of the samples on expo-
sure to air. It is hoped that these measures will allow quantitative iden-
tification of green rust in the recovered core samples.
Further research into this hazard is urgently required to identify
Green rust detection and analysis structural and lithological controls on the migration and trapping of
It is hypothesized that green rust is extremely widely distributed on hypoxic gas and to develop a rapid, reliable and quantitative tech-
the surface of sediment grains. Its rapid oxidation when in contact nique for the identification of green rust in fresh or properly pre-
with air means that it is absent in all but the freshest of cores and served samples. The authors welcome contact from drillers,
may require specialist sampling and preparation methods to identify engineers, geologists and anyone else with experience of encoun-
that it is present. However, it is considered here to be the most likely tering hypoxic conditions, pressurized anoxic gas, or elevation of
chemical cause of deoxygenated gas in central London. Furthermore, CO2 levels on site. Confidentiality will be respected.
the absence of green rust and the presence of brown rust in Upnor
Formation sediments from cores taken within the unsaturated, Lower Conclusions
Aquifer cone of depression indicate that large volumes of sediment
in central London may have been oxidized in the last c. 150years, This study has identified two key hazards for geotechnical engineer-
leaving behind a nitrogen-rich hypoxic gas, now pressurized by ing in Lambeth Group (particularly Upnor Formation) sediments in
recharge of that aquifer. London: the misidentification of the risk of confined space hypoxia
A full programme of laboratory testing on borehole samples recov- by reliance on observations of glauconite; and the likelihood of
ered during the ground investigation phase of the Thames Tideway encountering pressurized deoxygenated, nitrogen-rich gas following
Tunnel project is under way to identify the mineralogy of reduced and short- or long-term changes in groundwater level.
oxidized soil samples. The rapid oxidation of green rust means that its Glauconite should no longer be considered the most important
recovery from field samples for laboratory testing is extremely difficult. predictor for confined space hypoxia and the absence of glauconite
Samples recovered to date do contain glauconite and goethite (a possi- should not be taken to imply a lower risk of its incidence. Lambeth
ble oxidation product of green rust) but green rust itself has not yet been Group geology is complex and different factors appear to be impor-
quantified, other than being briefly visible during sample recovery. Part tant at different locations or stratigraphic depths but the research to
of the reason for this may be sample retrieval and preparation. At pre- date strongly suggests that other minerals, such as green rust, pyrite
sent the procedure is that immediately following recovery from the and organic carbon, have a much more important role in the occur-
ground and while still within a protective, rigid plastic lining, the rence of confined space hypoxia. Whereas pyrite and organic car-
100mm diameter core is cut into 100mm length samples at predeter- bon are relatively easy to identify and may leave a characteristic
mined intervals within the stratigraphic sequence. Liquid wax is poured signature following oxidation, to date green rust has proved par-
onto the open ends of the cut core to form an impenetrable barrier ticularly difficult to identify, even in samples collected specifically
against potential oxidizing effects and each sample is then wrapped in a for this research project. In very fresh samples (such as while drill-
layer of cling film. Plastic caps are placed on both ends and then the ing or boring) green rust can be identified by its characteristic blu-
whole sample is wrapped in a second layer of cling film, labelled and ish green colour, but it can be seen only within the first few minutes
placed within a clear plastic soil sample bag sealed with tape. This pro- of sample extraction as it oxidizes rapidly.
cedure is undertaken as rapidly as possible to minimize exposure of the Past and current water extraction practices pose significant risks
soil sample to oxidation. None the less, when the sample is unwrapped owing to the lowering and then rising groundwater leading to trapped
in a specially designed anaerobic chamber (glove box) a measurable deoxygenated air in sand beneath clay. The limited gas flow data col-
increase in oxygen levels is recorded, implying that the waxing and lected to date indicate that the permeability of gas in Upnor Formation
wrapping technique is not entirely effective in preventing oxygen from sediments is similar to that of water (i.e. c. 106ms1). Under these con-
becoming trapped in the wrapping of the sample. To overcome this ditions, a single borehole would take almost a year to relieve the gas
issue, future core samples will be vacuum-packed immediately upon pressure in soil out to a 30m radius, and thus it is most probably imprac-
core recovery and then frozen in an attempt to inhibit oxidation. tical to relieve the excess gas pressure encountered within an engineer-
At present the XRD (X-ray diffraction) and FTIR (Fourier trans- ing project. Although this may not be an immediate concern for
form infrared) analyses cannot be undertaken in an oxygen-free pressure-balancing tunnel boring machines (TBMs), oxygen-poor gas
environment, with the result that all samples are oxidized during test- may readily seep through some concrete tunnel linings, particularly dur-
ing. In future, samples will be mixed with glycerol, which acts a barrier ing the first year of installation (Pihlajavaara 1991), posing a hazard to
DEOXYGENATED GAS IN THE LAMBETH GROUP 177

personnel during fitting out or inspection work. Indeed, the high for- Jones, M.A. 2007. Rising groundwater in central London. Water and Sewerage
ward ground pressures induced by pressure-balancing TBMs may Journal, 4, 3536.
Legrand, L., Mazerolles, L. & Chausse, A. 2004. The oxidation of carbonate
exacerbate this problem.
green rust into ferric phases; solid-state reaction of transformation via solution.
Geochimica et Cosmochimica Acta, 68, 34973507.
Acknowledgements. The authors are particularly grateful to Lewis, J.L. & Harris, J.R. 1998. The engineering implications of deoxygenated
P. Gawne, P. Hemmings, G. Howells, U. Lawrence, A. Miller, gases in the Lambeth Group of north east London: A case history. CIRIA
C. Warren, E. Woods, A. Sefton and V. Yuan for their invalu- Research Project, 576, 11.
able insights and recollections of the events discussed in the Lucas, H.C. & Robinson, V.K. 1995. Modelling of rising groundwater levels in
the Chalk aquifer of the London Basin. Quarterly Journal of Engineering
text. Laboratory data were primarily provided by J. Huggett, J.
Geology, 28, S51S62.
Najorka and R. Court. The authors are indebted to J. Standing Mackenzie, F.T. 2005. Sediments, Diagenesis and Sedimentary Rock: Treatise on
and S. Shaw for helpful insights and guidance, and to A. Bolsher, Geochemistry. Elsevier, Oxford.
S. Ackerley and G. Keefe for technical support. Gratitude is also Marsh, T.J. & Davies, P.A. 1983. The decline and partial recovery of groundwa-
expressed to the referees of this paper, who made many useful ter levels below London. Proceedings of the ICE, 74, 263276.
suggestions and comments. The research was funded through the Morgan, H.D. & Bartlett, J.V. 1969. Tunnel design. The Victoria Line.
Proceedings of the ICE, Supplement Volume, 377395.
Thames Tideway Tunnel project (Thames Water Utilities Ltd)
Nagata, F., Inoue, K., Shinoda, K. & Suzuki, S. 2009. Characterisation of
under the direction of J. Harris and J. Greenwood. formation and oxidation of green rust (Cl) suspension. ISIJ International,
49, 17301735.
Newman, T.G. & Hadlow, N. 2011. Ground challenges for the Thames Tunnel.
References In: redmond, S. & romero, V. (eds) Proceedings of the Rapid Excavation
and Tunnelling Conference, San Francisco, 1922 June. Society for Mining,
Aldiss, D.T. 2012. The stratigraphical framework for the Palaeogene suc- Metallurgy and Exploration Inc, Colorado, 204218.
cessions of the London Basin, UK. British Geological Survey Open Report, Newman, T.G. & Wong, H.-Y. 2011. Sinking a jacked caisson within the
OR/12/004. World Wide Web Address: http://nora.nerc.ac.uk/17121/. London Basin geological sequence for the Thames Water Ring Main
Bentor, Y.K. & Kastner, M. 1965. Notes on the mineralogy and origin of glau- extension. Quarterly Journal of Engineering Geology and Hydrogeology,
conite. Journal of Sedimentary Petrology, 35, 155166. 44, 221232.
BSI. 2011. Code of practice for health and safety in tunnelling in the construction Newman, T.G., Yuan, L.-F.V. & OKeeffe, L.C. 2010. Using tunnel boring
industry. British Standards Institute, London. data to augment the geological model. Proceedings of the Institute of Civil
Christiansen, B.C., Balic-Zunic, T., Dideriksen, K. & Stipp, S.L.S. 2009. EngineersGeotechnical Engineering, 163, 157166.
Identification of green rust in groundwater. Environmental Science and Odin, G.S. & Matter, A. 1981. De glauconiarum origine. Sedimentology, 28,
Technology, 43, 34363441. 611641.
Clark, K.A. 1970. Discussion, on: FOLLENFANT et al. 1969. The Victoria OLoughlin, E.J., Larese-Casanova, P., Scherer, M. & Cook, R. 2007. Green
Line, Proceedings of the ICE, Supplement Volume, 328329. rust formation from the bioreduction of -FeOOH (lepidocrocite): Comparison
CNPlus. 1991. Thames Water and Davy Mining fined over lapse in safety of several Shewanella species. Geomicrobiology Journal, 24, 211230.
standards at Thames Ring Main Project in London. Construction News, 16 Page, D.P. & Skipper, J.A. 2000. Lithological characteristics of the Lambeth
May 1991. Group. Ground Engineering, 33, 3843.
Doyle, B.R. 2001. Hazardous Gases Underground: Applications to Tunnel Parmar, N., Gorby, Y.A., Beveridge, T.J. & Ferris, F.G. 2001. Formation of
Engineering. Marcel Dekker, New York. green rust and immobilization of nickel in response to bacterial reduction of
Ellison, R.A. 1983. Facies distribution in the Woolwich and Reading Beds of the hydrous ferric oxide. Geomicrobiology Journal, 18, 375385.
London Basin. Proceedings of the Geologists Association, 94, 311319. Pettijohn, F.J., Potter, P.E. & Siever, R. 1987. Sand and Sandstones. Springer,
Ellison, R.A., Knox, R.W.O.B., Jolley, D.W. & King, C. 1994. A revision New York.
of the lithostratigraphical classification of the early Palaeogene strata of the Pihlajavaara, S.E. 1991. Long-term gas permeability properties of con-
London Basin and East Anglia. Proceedings of the Geologists Association, crete in wet repository conditions. Nuclear Engineering and Design,
105, 187197. 129, 4148.
Ellison, R.A., Woods, M.A., Allen, D.J., Forster, A., Pharoah, T.C. & King, C. Refait, Ph., Benali, O., Abdelmoula, M. & Genin, J.-M.R. 2003. Formation of
2004. Geology of London. Memoir of the British Geological Survey. ferric green rust and/or ferrihydrite by fast oxidation of iron(IIIII) hydroxy-
Feder, F., Trolard, F., Klingelhofer, G. & Bourrie, G. 2005. In situ Mssbauer chloride green rust. Corrosion Science, 45, 24352449.
spectroscopy: Evidence for green rust (fougerite) in a gleysol and its miner- Simpson, B., Blower, T., Craig, R.N. & Wilkinson, W.B. 1989. The engineering
alogical transformations with time and depth. Geochimica et Cosmochimica implications of rising groundwater levels in the deep aquifer beneath London.
Acta, 69, 44634483. CIRIA Special Publication, 69.
Flavin, R.J. & Joseph, J.B. 1983. The hydrogeology of the Lee Valley and Taylor, R.M. & McKenzie, R.M. 1980. The influence of aluminum on iron
some effects on artificial recharge. Quarterly Journal of Engineering oxides. VI. The formation of Fe(II)Al(III) hydroxy-chlorides, -sulfates and
Geology, 16, 6582. -carbonates as new members of the pyroaurite group and their significance in
Genin, J.-M.R., Aissa, R., et al. 2005. Fougerite and FeIIIII hydroxycarbonate soils. Clays and Clay Minerals, 28, 179187.
green rust; ordering, deprotonation and/or cation substitution; Structure of Trolard, F., Bourrie, G., Abdelmoula, M., Refait, P. & Feder, F. 2007.
hydrotalcite-like compounds and mythic ferrosic hydroxide Fe(OH)(2+x). Solid Fougerite, a new mineral of the pyroauriteiowaite group: Description
State Sciences, 7, 545572. and crystal structure. Clays and Clay Minerals, 55, 323334.
Hansen, H.C.B. 1989. Composition, stabilization and light absorption of Fe (II) Wilkinson, B. 1985. Rising groundwater levels in London and possible
and Fe (III) hydroxycarbonate ('green rust'). Clay Minerals, 24, 663669. effects on engineering structures. In: downing, R. A. & jones, G. P.
HSE. 2009. Safe work in confined spaces: Confined Spaces Regulations 1997, (eds) Memoirs of the 18th Congress of the International Association of
Approved Code of Practice, Regulations and guidance, 2nd edn. Health and Hydrogeologists, Cambridge Eng, Cambridge, 145157.
Safety Executive, HMSO, Norwich. Zegeye, A., Ona-Nguema, G., Carteret, C., Huguet, L., Abdelmoula, M.
Huggett, J.M. & Knox, R.W.OB. 2006. Clay mineralogy of the Tertiary onshore & Jorand, F. 2005. Formation of hydroxysulphate green rust 2 as a sin-
and offshore strata of the British Isles. Clay Minerals, 41, 546. gle iron(IIIII) mineral in microbial culture. Geomicrobiology Journal,
Hurst, C.W. & Wilkinson, W.B. 1986. Rising groundwater levels in cit- 22, 389399.
ies. In: Cripps, J.C., Bell, F.G. & Culshaw, M.G. (eds) Groundwater in Zugibe, F.T., Costello, J.T., Breithaupt, M.K., Zappi, E. & Allyn, B. 1987.
Engineering Geology. Geological Society, London, Engineering Geology The confined space-hypoxia syndrome. Journal of Forensic Science, 32,
Special Publications, 3, 7580. 554560.

Received 16 February 2012; accepted 17 January 2013.


'HHU +RZLH DQG =XVVPDQ

5RFN)RUPLQJ 0LQHUDOV
)RU IXOO GHWDLOV VHH WKH 2QOLQH %RRNVKRS ZZZJHROVRFRUJXNERRNVKRS

5RFN)RUPLQJ 0LQHUDOV 9RO $ 0LFDV 6HFRQG (GLWLRQ


%\ 0( )OHHW
,Q WKLV H[WHQVLYH DQG FRPSUHKHQVLYH UHYLHZ WKH PLFD VHFWLRQV RI 9ROXPH  KDYH EHHQ FRPSOHWHO\ UHZULWWHQ UHRUJDQL]HG DQG JUHDWO\
H[SDQGHG 7KH WH[W FRYHUV VXFK DVSHFWV DV FU\VWDO VWUXFWXUH FU\VWDO FKHPLVWU\ FKHPLFDO VSHFWURVFRS\ VWDELOLW\ DQG SKDVH UHODWLRQV
UHGR[ EHKDYLRXU DQG ZHDWKHULQJ VWDEOH LVRWRSH JHRFKHPLVWU\ JHRFKURQRORJ\ DQG LJQHRXV DQG PHWDPRUSKLF JHRFKHPLVWU\ DQG
SHWURORJ\
,6%1  _ +DUGEDFN _  SDJHV _ 3XEOLFDWLRQ GDWH  )HEUXDU\ 
/LVW SULFH e )HOORZ
V SULFH e 2WKHU VRFLHWLHV SULFH e 2QOLQH %RRNVKRS &RGH 5)0$

5RFN)RUPLQJ 0LQHUDOV 9RO % /D\HUHG 6LOLFDWHV ([FOXGLQJ 0LFDV DQG &OD\ 0LQHUDOV
6HFRQG (GLWLRQ
%\ :$ 'HHU 5$ +RZLH - =XVVPDQ
7KLV H[WHQVLYH UHYLVLRQ GHDOV ZLWK WKH PLQHUDOV WDOF S\URSK\OOLWH FKORULWH VHUSHQWLQH VWLOSQRPHODQH ]XVVPDQLWH SUHKQLWH DQG
DSRSK\OOLWH 7KH WH[W KDV EHHQ FRPSOHWHO\ UHZULWWHQ DQG YHU\ PXFK H[SDQGHG WR WDNH DFFRXQW RI WKH PDQ\ DGYDQFHV WKDW KDYH EHHQ
PDGH LQ DOO DVSHFWV RI WKH (DUWK VFLHQFHV QRW OHDVW PLQHUDORJ\
,6%1   _ +DUGEDFN _  SDJHV _ 3XEOLFDWLRQ GDWH  0DUFK 
/LVW SULFH e )HOORZ
V SULFH e 2WKHU VRFLHWLHV SULFH e 2QOLQH %RRNVKRS &RGH 5)0%

5RFN)RUPLQJ 0LQHUDOV 9RO % )UDPHZRUN 6LOLFDWHV 6HFRQG (GLWLRQ


%\ :$ 'HHU 5$ +RZLH :6 :LVH DQG - =XVVPDQ
7KLV PDMRU UHYLVLRQ WDNHV SODFH  \HDUV DIWHU SXEOLFDWLRQ RI WKH UVW HGLWLRQ DQG GHDOV ZLWK IHOGVSDWKRLGV VLOLFD PLQHUDOV DQG ]HROLWHV
7KH WH[W KDV EHHQ FRPSOHWHO\ UHZULWWHQ DQG YHU\ PXFK H[SDQGHG LQFRUSRUDWLQJ WKH DGYDQFHV LQ NQRZOHGJH DQG XQGHUVWDQGLQJ DULVLQJ
IURP WKH QHZ DQG LPSURYHG WHFKQLTXHV IRU WKH VWXG\ RI PLQHUDOV WKDW KDYH GHYHORSHG
,6%1  _ +DUGEDFN _  SDJHV _ 3XEOLFDWLRQ GDWH  0D\ 
/LVW SULFH e )HOORZ
V SULFH e 2WKHU VRFLHWLHV SULFH e 2QOLQH %RRNVKRS &RGH 5)0%

5RFN)RUPLQJ 0LQHUDOV 9RO $ 1RQ6LOLFDWHV 2[LGHV +\GUR[LGHV DQG 6XOSKLGHV


6HFRQG (GLWLRQ
%\ -): %RZOHV 5$ +RZLH '- 9DXJKDQ DQG - =XVVPDQ
$ FRPSOHWHO\ UHZULWWHQ DQG PXFK H[SDQGHG QHZ HGLWLRQ RI WKH SDUW RI 9ROXPH  RI 5RFN)RUPLQJ 0LQHUDOV GHDOLQJ ZLWK WKH
R[LGH K\GUR[LGH DQG VXOSKLGH PLQHUDOV ,W LV DQ HVVHQWLDO UHIHUHQFH ZRUN IRU UHVHDUFKHUV WHDFKHUV DQG DGYDQFHG VWXGHQWV RI
PLQHUDORJ\ SHWURORJ\ JHRFKHPLVWU\ PLQLQJ JHPPRORJ\ DQG UHODWHG HOGV
,6%1   _ +DUGEDFN _ 3DJHV  _ 3XEOLFDWLRQ GDWH  -XQH 
/LVW SULFH e )HOORZ
V SULFH e 2WKHU VRFLHWLHV SULFH e 2QOLQH %RRNVKRS &RGH 5)0$

2UGHU IURP 7KH *HRORJLFDO 6RFLHW\V /\HOO &ROOHFWLRQ MRXUQDOV 6SHFLDO 3XEOLFDWLRQV DQG
ZZZJHROVRFRUJXNERRNVKRS
Quarterly Journal of Engineering Geology and Hydrogeology, Vol. 46, 2013, pp. 167177
doi: 10.1144/qjegh2012-013
2013 The Geological Society of London
ERRNV RQOLQH )RU PRUH LQIRUPDWLRQ YLVLW ZZZJHROVRFRUJXN/\HOO&ROOHFWLRQ
Published Online First on March 25, 2013

You might also like