Density

The essential physical characteristic of the liquid is its density. The density of the liquid, in conjunction with the liquid column height, determines the pressure exerted by the liquid on the borehole ice surface. This pressure minimizes the abrupt isostatic pressure release during the drilling process, which in turn minimizes stress-related lateral fractures produced in the ice core and thus improves ice-core quality. This drilling issue is first apparent at depths of ∼100 m where normally the drill liquid is first employed. The density is also essential to produce a balanced even distribution of pressure against the borehole wall to minimize any long-term borehole deformation.

For intermediate to deep drilling operations there are two options to match the drill liquid pressure to the ice isostatic pressure. Adjusting the liquid head height is by far the simpler and can give good pressure matching at specific depth ranges, though not necessarily the whole depth. When the liquid column height is used in combination with adjusting the density (e.g. by adjusting the mixing ratio of two liquids), the pressure matching can be set not only more precisely but over a much wider depth and temperature range.

The ice isostatic pressure is related to the ice density, which in turn is related to the ice temperature and depth and the weight of ice pressing down from above. For many sites in Greenland and the Antarctic the ice density below the firn varies from 910 to 925 kg m⁻³. For practical purposes a density of ∼920–950 kg m⁻³ (Reference Gundestrup, Clausen, Hansen and JohnsenGundestrup and others, 1994) over a temperature range 0°C to −59°C is desirable, and preferably with the ability by mixing to fine-tune by ±25 kg m⁻³.

Viscosity

The majority of drill systems used in ice-core drilling operations are wire line types. The drill is lowered down the borehole from the surface using a cable, a core is drilled, the drill and core are hoisted to the surface, the ice core is removed, the drill is cleaned and the process is repeated. This incremental drilling operation directly affects the duration of a complete drilling operation in the very expensive working environment, which is usually limited to a 1–3 month long drilling season. Thus the duration of the drill runs has a significant influence on the overall project costs. A slow operation might add 1 year or more to a project compared with a faster operation.

A significant contributor to the overall drilling operation is the time taken for the drill to descend and ascend the borehole (Reference Vasiliev and KudryashovVasilievand Kudryashov, 2002) particularly when recovering ice core at greater depths such as at EPICA Dome C 2 where at 3200 m 3 hours were required to recover 40 cm of ice core. The drill travel speed is related to the viscosity of the drill liquid. To allow the drill to descend and ascend quickly, a low kinematic viscosity is a desirable criterion: preferably below 25 mm² s⁻¹ (Reference Popp, Hansen, Sheldon and PantonSheldon and others, 2014) for a NEEM-configured drill, or 5 mm² s⁻¹ (Reference Gundestrup, Clausen, Hansen and JohnsenGunderstr-up and others, 1994) for an ISTUK configuration.

Compatibility

For ice cores to be scientifically useful it is essential that the drill liquid has either no or very little effect on the ice or on any of the scientifically interesting chemical species contained within the core. This liquid should also be compatible with materials used in and around the drill site, the logging process and storage facilities. It should be compatible with substances or techniques used during drill emergency recovery situations and with modern borehole logging probes.

Further essential properties

1. Low toxicity is essential for the work environment and to minimize the environment impact

2. Availability, both present and in the foreseeable future

3. Easy logistical transport

4. Stability in the environmental conditions and over time

5. Overall costs

Physical properties: density

From 2011 to 2013 we carried out several density measurements on the candidate liquids from EstiChem A/S. In the case of ESTISOL™ 140, two batches were tested giving similar results and following the published data of Reference TalalayTalalay and others (2014) (Table 4; Fig. 1).

Fig. 1.

Density temperature characteristics: comparison of ESTISOL™ 140, ESTISOL™ 165 and ESTISOL™ 240/COASOL™ mixture. For a range of temperatures from 0°C to −35°C a mixture of ESTISOL™ 140 and ESTISOL™ 165 from ∼17% to 0% appears to fit into the usable density band. For temperatures from −36°C to −66°C pure ESTISOL™ 140 appears to fit into the usable density band.

Table 4.

Density characteristics for the two candidate liquids at various temperatures

Physical properties: viscosity

The dynamic viscosity of the candidate drill liquids was measured for the two batches of ESTISOL™ 140 and for ESTISOL™ 165, and with the measurements of density the kinematic viscosity was determined. Some variation was found between the batches of ESTISOL™ 140 from 2011 and 2013, and compared with the published data of Reference TalalayTalalay and others (2014) (Table 5; Fig. 2).

Fig. 2.

Kinematic viscosity temperature characteristics: comparison of ESTISOL™ 140, ESTISOL™ 165 and ESTISOL™ 240/COASOL™ mixture. For temperatures down to −51°C ESTISOL™ 140 appears to have a kinematic viscosity similar to or lower than what was experienced at NEEM. Using the kinematic viscosity limit of 35 mm² s⁻¹ gives a working temperature limit for ESTISOL™ 140 of −54°C from our measurements, similar to the results of Reference TalalayTalalay and others (2014).

Table 5.

Key values of dynamic and kinematic viscosity at room temperature to −66°C

A drill liquid with low kinematic viscosity is desirable for reducing transit times and ascent system stresses. (A consequence of using a liquid with high kinematic viscosity is an increase in resistance to drill movement. During drill descent, where gravity is employed, this decreases the descent velocity. The drill ascent is achieved using cable tension exerted to overcome the force of gravity, to displace the liquid within the tight confines of the borehole and to overcome the relatively small frictional forces. Keeping the drilling parameters constant (e.g. drilling depth, borehole diameter, ascent speed, ice-core length), an increase in kinematic viscosity will require a greater cable tension to displace the liquid. This increase in tension increases stresses within the hoisted drill, the cable, the hoisting system and the power supply generators. This in turn increases component wear and fatigue, overall fuel consumption, possible component weight and associated project costs.) At NEEM we successfully used a drill liquid combination of ESTISOL™ 240 and COASOL™ with a kinematic viscosity of 25 mm² s⁻¹ for the first 5 0% of the drilling, which then slowly reduced to 7.5 mm² s⁻¹ at the ice-sheet base (Reference Popp, Hansen, Sheldon and PantonPopp and others, 2014b). By extrapolating from these experiences we predicted for our NEEM drill configuration a reasonable working upper limit of kinematic viscosity of ∼35 mm² s⁻¹ (Reference Popp, Hansen, Sheldon and PantonSheldon and others, 2014). From our density and viscosity measurements, and extrapolating the 35 mm² s⁻¹ upper limit, we predict that ESTISOL™ 140 and ESTISOL™ 165 will be an effective drill liquid for temperatures ranging from 0°C to −54°C.

Hydrophobic properties

The liquids were tested for hydrophobic properties, initially by mixing with water at room temperature and then with small granules of ice at −16°C, −25°C and −55°C.

At room temperature the water and candidate liquids were poured into a glass graduated cylinder and mixed thoroughly by sealing the cylinder top and shaking. The mixed components rapidly separated into two distinct layers: the water in the lower layer, the ESTISOL™ 140 in the upper layer; and the water in the upper layer, the ESTISOL™ 165 in the lower layer. The liquids were repeatedly mixed by either stirring or shaking. On each occasion they rapidly separated into well-defined visible bands.

In a −16°C freezer, granules of ice were poured into, and then mixed by vigorous shaking with, cooled ESTISOL™ 140, ESTISOL™ 165 and a mixture of both containing 17% (vol.) ESTISOL™ 165. With pure ESTISOL™ 140 the ice granules were suspended for some time, drifting only very slowly to the lower section of the container. With pure ESTISOL™ 165 the ice granules floated rapidly to the upper area of the container. With the ESTISOL™ 140/ESTISOL™ 165 mixture the majority of the ice had formed in the upper area within ∼1 min, forming a stable well-defined visible separation of ice chips and drill liquid. The process was repeated, with similar observations. The mixtures were left for several days without any noticeable change and were subsequently left for a total of 2 years without any noticeable change to the separated ice and drill liquid, or any visible dissolving of ice.

The test procedures were repeated with a second batch of samples in the −25°C freezer. The observations from the tests carried out at −25°C were very similar to the tests performed at −16°C with the exception that the mixture of pure ESTISOL™ 140 with ice separated very slowly; the ice was almost suspended. These mixtures were left for a total of 2 years without any noticeable change to the separated ice and drill liquid, or any visible dissolving of ice.

Further tests were carried out in a chest freezer at −55°C, similar to the tests performed at −16°C and −25°C. For all the mixtures in these tests the ice granules were observed to float to the surface of the containers.

No noticeable effects

1. High-density blown film polyethylene (ice-core bags)

2. Polyethylene film (bubble wrap)

3. Polypropylene (PP)

4. Polytetrafluoroethylene (PTFE/Teflon)

5. Polyesters (in fabrics, bottles, screens, etc.)

6. Rigid polystyrene

7. Water and ice – hydrophobic/non-polar

8. Wood, paper, cardboard, paper packaging

9. Metals (Al, Fe, Ti, Cr, Au)

Partially affected

Fig. 4.

O-ring after immersion for 2 days (right); O-ring with no immersion (left).

Fig. 5.

O-ring returns to approximately original dimensions after 7 days at room temperature (right); O-ring with no immersion (left).

Affected

EPS foam: dissolved within seconds

The effects on EPS are of particular interest to the ice-core community as some types of ice-core box used for storing and transporting ice cores are made of EPS. After experiencing similar effects using ESTISOL™ 240 at Flade Isblink in 2006, we changed the composition of the Danish ice-core boxes to expanded polypropylene (EPP). These ice-core boxes were tested and appeared to show no degrading effects.

Fig. 6.

No noticeable effect on boot soles after 2 weeks of exposure to ESTISOL™ 140 at ABN.

Ice cores

Within 8 days of drilling in both dry and wet modes, the tests were concluded at a depth of 131 m. When the ice cores were removed from the drill they were wiped dry using paper towels, similar to our previous operations at Flade Isblink (2006) and NEEM (2009–10 and 2011) (Fig. 7). This removed most of the excess liquid still on the ice-core surface. The cores were then analysed using the dielectric profiling (DEP) method before packing into ice-core boxes for shipping to Copenhagen, Denmark. In Copenhagen a very slight insignificant odour could be discerned during the unpacking. No discernible odour was noticed during electrical conductivity measurement (ECM) analysis of the ice core. Generally, we can report that ESTISOL™ 140 acted as a good drill liquid, eliminating lateral fractures, producing consistently excellent quality ice core and functioning well as an intermediate-depth drill liquid.

Fig. 7.

First ice cores to be drilled using ESTISOL™ 140 giving consistently good-quality ice cores.

Drill handling

Our experience of handling the drill covered with ESTISOL™ 140 at −29°C is similar to that with previous drill liquids. To reduce discomfort for the drill operator, liquid-resistant apparel is a necessity. Using this liquid allowed clothing to be easily dried in a well-ventilated space at 35°C. With surface air temperatures ranging from −10°C to −25°C, a noticeable odour similar to fly repellent could be detected. To reduce this, we recommend use of good ventilation within all drilling, core-logging, ice-core-handling and clothes-drying areas.

Ice-cutting chips

From our experience an important factor in a successful ice-core drilling operation is the size and consistency of the ice-cutting chips. With a low-viscosity drill liquid, chips that are either too fine or too large will not be transported into the drill’s chip chamber and will lead to clogging of the drill’s internal chip transport cavity between the core barrel and the outer core barrel. This can lead to reduced ice-core run lengths, lost chips in the borehole and generally bad chip recovery. Using our HT drill with a ‘normal’ set-up pitch of 4.0 mm with ESTISOL™ 140 produced ice chips with ‘good’ granular consistency which gave good chip transportation (Fig. 8). Ice-core lengths were an average of 1.53 m, chips were found mainly in the chip chamber and no chips were found on the top of the core.

Fig. 8.

Drill chips collected at 125 m: good chip size (∼2–3 mm) and recovery.

Drill liquid/ice separation using liquid phase fractionation

The ice-cutting chips from the HT drill chip chamber were collected (Fig. 9) and left to melt overnight in a room at ∼35°C. The water and ESTISOL™ 140 separated into two very clear bands: water in the lower band with ∼65% by volume and ESTISOL™ 140 in the upper remaining band. The mixture was then left outside to cool and the water to refreeze. The ESTISOL™ 140 could then be easily and effectively removed without the ice (Fig. 10), giving a ‘proof of concept’ of a liquid phase fractionating method for separating the water and drill liquid to allow the reuse of the ESTISOL™ 140. The refrozen water was then allowed to melt again. Only a thin film of drill liquid could be seen, ∼1% or less by volume. From our experience this compares with ∼7–10% loss when using the more traditional ‘centrifuge method’ used at NorthGRIP and NEEM.

Fig. 9.

Ice chips before melting.

Fig. 10.

Ice chips after melting and refreezing.

Measured drill ascent and descent speeds

In a 126 mm diameter borehole the drill descent speed was ∼1.0 m s⁻¹ and the ascent speed was ∼0.8 m s⁻¹ (full drill with ∼2500 N tension).

The low kinematic viscosity of ESTISOL™ 140 ensured relatively fast transit times for the drill and reduced system stresses during the hoisting procedure. Drill descent speeds of 1.0 m s⁻¹ were achieved within the 126 mm diameter borehole at temperatures of −29°C, implying a drill liquid kinematic viscosity of 3 ± 1 mm² s⁻¹. This is 30% faster than the speed (0.7 m s⁻¹) achieved with a similar HT drill set-up using ESTISOL™ 240 at a higher temperature of −15.7°C with a kinematic viscosity of 13.5 ± 1 mm² s⁻¹.

Summary of NEEM 2012 field tests

1. Drilling and recovery of ice core went well and without issues.

2. Recovered consistently excellent quality ice cores

3. No problems were encountered cleaning the ice cores. As with our previous experience with other drill liquids, we recommend thorough cleaning of the ice cores in the field before packing

4. Recovery of liquid from the chips is very good

5. Excess liquid splashed on the driller’s clothing, boots and gloves dried readily at 35°C

6. With surface temperatures of −10°C to −25°C a noticeable odour similar to fly repellent could be discerned; therefore, at these temperatures, ventilation is recommended

7. Drill descent speeds of 1.0 m s⁻¹ were achieved at temperatures of −29°C, 30% faster than using ESTISOL™ 240 at −15.7°C

8. ESTISOL™ 140 appears to be a candidate for drilling operations at ice temperatures well below –29°C

Drilling performance at ABN

Systematic investigation into ice-core quality

During the drilling operation the recovery of high-quality ice core was a prime objective and to this end a systematic quality control was put in place. This involved (1) checks for possible missing core using ice-core break matching techniques during logging and comparing drill depth with logging depth for any unexplained differences, (2) visual checks for external cracks, scratches or damage and (3) visual checks for internal fractures and cracks.

From previous experience of using a drill liquid at sites where melt layers are minimal, very small internal cracks can develop at depths of 200 m and below if the liquid column is not high enough (Popp and Hansen, unpublished information). At ABN we measured the liquid column height with each drilling run and gave attention to whether any very small internal fractures were starting to develop in the ice core. When small fractures started to appear, the liquid column was immediately increased in height by pouring more ESTISOL™ 140 into the borehole. From this work we quantified the absolute minimum liquid column required to preserve the recovery of good-quality ice core. Later, we compared these data to the calculated isostatic pressure of the ice (Fig. 12). We found that if the difference between the hydrostatic pressure exerted by the drill liquid column and the ice isostatic pressure was >1.75 MPa, very small internal fractures would start to occur.

Fig. 12.

Systematic ice-core quality control. Red line indicates the absolute minimum liquid column height at a given drill depth. Dashed blue line indicates the liquid column height equivalent of the calculated ice isostatic pressure.

We extrapolated these data and so calculated the liquid column height required to retain high ice-core quality at greater depth. We estimated that the liquid level (i.e. the depth from the present-day snow surface to the top of the liquid column) should ideally be kept at 164 m; it should not be allowed to fall below 220 m (Fig. 13).

Fig. 13.

Estimates of the drill liquid column height required to ensure good ice-core quality.

From this extrapolation we estimated the total drill liquid requirement for a given future project. Assuming 2 5% loss of drill liquid (past experience shows ∼16% loss), Table 6 shows the calculated quantities of drill liquid to drill to ≥500 m.

Table 6.

Calculated quantities of drill liquid required for a 129.6 mm diameter borehole to retain good ice-core quality for various intermediate-depth drilling projects

Drill liquid recovery

An essential procedure of a drilling operation is the efficient recovery and subsequent reuse of the drill liquid brought up to the surface inside the drill. With each run many litres of drill liquid are pulled out of the borehole as either almost pure liquid trapped in the closed spaces inside the drill or as a chip mixture of the ice cuttings and drill liquid. During the NEEM 2012 field tests, ∼35% of the chip mixture was found to be ESTISOL™ 140, representing 2 L of drill liquid per metre of ice core recovered. A further 5.5 L of drill liquid per metre of ice core was found inside the drill. This amount of drill liquid (7.5 L per meter of recovered ice core) represents a possible loss of 1000 L (5 × 200 L drums) for every 133 m of ice core drilled.

Instead of employing a traditional centrifuge system, which normally weighs >150 kg and is not as efficient at liquid retrieval, at ABN we planned to use a liquid phase fractionating method to separate the ice from the drill liquid, similar to the concept tested at NEEM in 2012. We built a system consisting of a 200 L drum mounted horizontally with two 1.5 kW drum heaters surrounded by insulation padding. An opening was made on the upper portion of the drum to allow the mixture to be poured in. A tap was added to the lower part of the drum to allow the liquids to be removed from the drum, first water followed by ESTISOL™ 140.

During the whole drilling operation, 1900 L of fresh pure ESTISOL™ 140 was poured into the borehole. Using the liquid phase fractionating method ∼1266 L was recovered from the chips and slush brought to the surface by the drill. Of this amount, ∼766 L was poured back into the borehole and 500 L left over on the surface (Table 7).

Table 7.

Of the 1900 L ESTISOL™ 140 poured into the borehole, >1615 L was left over within the borehole or on the surface. Approximately 15% of the drill liquid was unaccounted for during the operation, probably due to a combination of seepage into the firn during the hoisting process, small surface spillages and errors in our measurements