DFDP-2 drilling started

Rupert Sutherland, Principal Investigator, GNS Science
John Townend, Principal Investigator, Victoria University of Wellington
Virginia Toy, Principal Investigator, University of Otago
Simon Cox, Quaternary Team Leader, GNS Science

The start of DFDP-2 drilling activities in the Whataroa Valley 1/9/2014.

The aim of the DFDP-2 project is to examine a major fault before it ruptures in a large earthquake. What are the ambient conditions in the fault zone? What are the physical properties of the materials? What is the fault zone’s internal structure? Answers to these and many other questions will help us understand what happens when a fault ruptures in an earthquake and how slip accumulates to form a plate boundary. Have a look at our video for an introduction to the DFDP-2 project:

Our results will have implications for global earthquake science. This is why scientists from twelve countries have come together to gather information on fault geometry, temperature and pressure conditions, collect rock and fluid samples, and install long-term monitoring equipment. It is also why the international community, via the International Continental Scientific Drilling Program (ICDP) has funded more than half of the project, and why we have been successful in obtaining funding from the Marsden Fund.
The DFDP-2 project will have specific value to New Zealanders. It is surely important to understand the largest source of seismic hazard in South Island. What will happen during and after the next major earthquake? Does the hazard vary from day to day? Does anything unusual happen before an earthquake? Could we develop a warning system? Does the Alpine Fault offer benefits as well as hazards? For example, does the fault host a geothermal resource? Our project will help address these questions too.

DFDP-2A during coring of sediments 17/9/2014.

Phase 1 of DFDP-2 involves drilling and installation of casing though sediments in the Whataroa Valley. Drilling started on 29 August 2014 with installation of 12” casing using a dual-rotary-air method. The 12” casing was advanced to 29 m depth, and then 10” casing to 125 m. At this point, we opted to mobilise the coring rig earlier than originally planned, to sample the sediments and determine how to proceed. The sediment layers proved thicker than we had thought, and sediment cores could provide valuable information.

On 14 September, we started coring sediments. After moderate recovery between 125 and 160 m, we drilled open-hole with a stratopack bit and rapidly progressed to 206 m. On 17 September, we stopped coring glacial diamictites at 212 m depth.

Rupert Sutherland and Simon Cox inspect core.

The project has already produced results that have surprised many and will lead to a new understanding of activity on the fault and the mechanism by which glaciers cut and fill valleys.
The relatively young (<500 yr) cobbly river gravels are only a few metres thick. Below this, to a depth of about 50 m, there is a sequence of pebbly river gravels that are at least 12,000 yr old. Deposits from an ancient river delta are found between 50 and 77 m depth, and below this, we discovered a thick sequence of silts, likely deposited on the bed of an ancient lake that filled the space carved out by a glacier. Below 197 m, we find evidence for rocks that were rafted and dropped into a fairly deep lake from floating ice as the glacier retreated.
We infer that the lake-bed was at least 300 m below the sea level of the time, which itself was at least 100 m lower than today. We are examining samples to see if it was a marine fiord. We have recovered many wood samples suitable for radio-carbon dating the thick sequence of sediments.

The steepness of the valley wall is a surprise to every scientist involved. There are very few slopes >50 m in extent that dip >45° in the modern topography. How has the rock face stayed so steep? Does this have implications for the strength of shaking during Alpine Fault earthquakes? We are only 1 km from the surface trace of the fault.
The static water level in the upper gravel unit (29 m depth) is about 5 m below ground surface. Below 120 m, we encountered artesian pressures, as anticipated, but a shut-in pressure could not be determined. Low flow rates (<0.3 l/min) imply low permeability, consistent with high silt contents.
We measured the temperature at 120 m depth in DFDP-2A and calculated a geothermal gradient of 139°C/km. This result needs to be interpreted with caution. First, slight artesian flow of warm water from the open hole below 120 m depth may be significant. The corrected value may be closer to 80°C/km. Second, we anticipated an elevated geothermal gradient and artesian pressures based on hydrogeological modelling, and it may not persist to greater depth.
Scientific drilling often provides unexpected results. The new sediment thickness, temperature, and fluid pressure data are useful for planning deeper operations, but we also have to now re-examine models of landscape evolution, geothermal circulation, and ancient fault movements in the Southern Alps.

Operations continue and the project is ramping up to its main phase. The first hole, DFDP-2A, is now a monitoring hole that we had originally intended to drill later. On 24 September, we began pulling casing and then we started a new borehole, DFDP-2B, about 10 m closer to the valley wall. We have since installed 16” casing to 47 m using dual-rotary-air and are about to start dual-rotary-mud. This has caused about a week’s delay, but we may make up time if bedrock is not much below 220 m. Stay tuned!

Finishing off - and next time we will go deeper

Our project has now moved into completion and demobilisation mode. We drilled boreholes DFDP-1A to 100.6 m and DFDP-1B to 151.4 m. The cores we collected have been carefully packed and taken back to the University of Otago in Dunedin, and the processing tents and most of the people are now also gone.

Wireline logging took several days because of borehole instability issues and yet more technical obstacles. We did not get clear runs through the whole borehole on some of the tools, but we have collected a very nice dataset with a wide range of tools: natural gamma, calliper, electrical resistivity, spontaneous potential, density,
porosity, full waveform sonic, dipmeter, and induction flow meter.

Installation of observatory equipment in borehole DFDP-1B is still under way. We successfully installed a 2 Hz seismometer at the base of the borehole. We have 4 piezometers in place to monitor pore fluid pressure, and will install a fifth tomorrow. Twenty five sensors monitor the average temperature gradient in the borehole and detailed thermal structure immediately above the fault. We will continue with observatory installation and testing tomorrow.

First photo: Full waveform sonic tool and wireline logging truck




Right: Observatory installation team: Jennifer Eccles, Mike Hasting, Simon Cox, myself (kneeling), Alex, John
Townend, and Jeremy Cole-Baker.


An initial observation that I found interesting was confirmation that fault gouge is an effective hydraulic seal. A preliminary estimate of pore pressure difference across the fault in DFDP-1B is equivalent to 33 m of water head. This partly explains borehole instability when the hole was left open: there was a “waterfall” within the borehole trying to wash fractured rocks above the fault into and down the borehole. When trying to clear the borehole in DFDP-1A, to grout a cement-bentonite seal into the fault, we punctured a temporary breach with a long plastic pipe.


There was the most extraordinary slurping noise as water was sucked out of the drill string into the lower pressure gravel aquifer beneath.


Egor lowers casing with a shoe bit and casing advancer inside


Over the next few days we will wind up field operations, demobilise equipment, and secure the permanent observatory. It is gratifying to know that, despite substantial challenges, our long hours, good planning, and excellent contractors have led to success for all of our main technical objectives. We collected lovely cores across the fault, ran a full suite of wireline logs, and installed a fault zone observatory. I specially thank the students involved and our visitors from Liverpool University. Everyone worked long hours and behaved in a very professional and collegial way.

The whole operation was very slick. You really would not have known what a steep learning curve we
had actually all been on. Finally, I give massive thanks to Egor our “senior” driller. He worked hard and long hours, never had a day off, solved the many technical challenges we created for him, and mostly kept a smile on his face too (though he didn’t like torrential rain).

This is Rupert signing out for now. Over following months the massive quantity of samples and data we collected will need to be carefully analysed and written up, so it may take some time before I poke my head up above water again. The next phase, DFDP-2 (>1 km), already has advanced plans, but we will certainly want to improve them now that we have had this practical experience and learnt so much about what we might find at greater depth.

Egor and Alan





Finally, in case you haven't seen it, here is a short introduction to the Alpine Fault Drilling project:

Alpine Fault drilled through again!

The customised system of open-hole drilling devised by Alan Speight (our head driller) worked fantastically. We rapidly progressed from 45 m to 93 m in 18 hours, including one unsuccessful attempt to core. We  changed back to coring again at 93 m because we noticed an obvious change in cuttings and mud colour.

A sample of cemented cataclasite is carefully extracted from the core catcher

Coring progressed smoothly this time, even though one of our drillers walked out on us for no apparent reason. For the last two days I have been working as a drillers offsider, trying to keep us on track with the project, but it has been tough to balance every demand.A sample of cemented cataclasite is carefully extracted from the core catcher.

Drama continued again last night when a sudden storm passed through at 2-4 a.m., flattening our core processing and logging facility. Fortunately, nothing catastrophic was lost or damaged.We rebuilt better and were ready to drill again by 7 a.m.

Cores of the Alpine Fault footwall emerge for the first time

Skies cleared during the morning, lovely cores continued to emerge from the borehole, and life was looking good. At around 4 p.m. today we ran into a slight problem: we hit an obstacle, obtained no core, and could not advance any further. We tripped out all the drill rods and checked the bit. It confirmed the suspicion that I had after inspecting the core catcher from our previous run. Light grey clay gouge was blocking mud circulation. Was it just a gouge-rich zone within the thick (>30 m) cataclasite zone, or was it the main fault?

At about 7 p.m. we collected the first core from beneath the problem zone (thanks to the skill of our driller on the levers, “Egor”). This core and the subsequent two confirmed that we had just crossed the principal slip surface of the Alpine Fault at a depth of 128 m. Fantastic!



Both of our boreholes have obtained high-quality cores across the fault, including the precious fault gouge of the primary fault surface. There is a clear sense of excitement and achievement, and a big group has decided to work through the night to see what is going to appear.The rock cores currently emerging are really interesting, because the rock types immediately beneath the fault are not well known, in comparison to the well exposed and studied sequence above.

A beautiful core of the Alpine Fault from 128 m

It is premature for me to say what the rocks are that we cored, but brilliant white and green shades indicated that we had crossed the fault. We decided to keep going for another  shift, and over the next 24 hours a preliminary geological description will be constructed.

If all goes to plan, we will have finished our main phase of drilling tomorrow. We will move onto
wireline logging and geophysical observatory installation. I personally now need a few hours
sleep after the drama of the previous two nights.

John Townend, Richard Norris, Rupert Sutherland, and Virginia Toy consider achievement so far

Drill Bits and Core Pieces

Understanding drilling technology and learning how it would handle the rocks of the Alpine Fault was a primary objective of phase 1 of our project. We definitely have been on a steep learning curve over the last few days.

We successfully grouted the lost casing shoe in place and cut through it, but we have now used quite a number of different steel casings. The outer casing (with lost shoe) has an internal diameter (ID) of 150 mm and reached 34 m. We then installed a

our PQ drill string from rattling about. We then inserted our drill string with core barrel and advanced the casing as we cored down. However, even after cutting away the lost shoe, it was very slow progress and veryPWT casing with ID 130 mm, to stop

poor core. By noon today we had still only reached a depth of 45 m and had little to show for it. We decided to abandon coring until we were closer to the fault.

First photo: Diamond coring bit and reamer.


Second photo : Poor core recovery with many fragments of collapsed borehole in the core barrel and a very short run of intact rock. Rates of drilling were less than 1 m per hour and this was one of the best cores collected.

In the meantime work continued on instruments to be put down the borehole, and people caught up on a bit of sleep. The rain was heavy at times and some were probably (secretly) a little glad that we did not have to work in a wet tent through the night.

Photo 3: Mike Hasting splices wires then shrink wraps a temperature sensor onto one of the sensor arrays. One array will measure thermal gradient within the well, while the other records temperature at a finer scale in the 10 m immediately above the fault.

We replaced the coring bit with a flat bit, in an attempt to more rapidly create an open hole. Unfortunately, this didn’t seem to work. It seemed that the borehole wall was causing us problems, providing fragments that were jamming against the drill rods. Very frustrating. Alan Speight, our head driller, then suggested that we try to increase the size of the annulus, so that cuttings had more space to escape. He came up with the

cunning plan of attaching the same size bit onto thinner drill rods, but welded some spacers to the rods, to keep them central and hence the hole straight. Finally, this seemed to be working. Between 10 pm and 11 pm we drilled 4 m.

Right. A cutting bit designed to create an open hole rather than collect core

Below. A cutting bit designed to create an open hole the same size (123 mm) as a PQ coring system, but the bit is attached to drill rods used with the smaller HQ coring system. Steel bars welded in place keep the rods central within the hole.

Close up to the Fault

Check out TVNZ's Close up coverage of the Alpine Fault Drilling from Thursday 27th January to get the picture!

Second drill goes down

After the fantastic high of perfectly coring the Alpine Fault on Sunday, We started to focus our attention on our second drill hole which we call DFDP 1-B.  

By Tuesday we saw the establishment of site DFDP-1B and installation of 6” steel casing. On Wednesday we continued to install casing, but had a failure and detachment of the casing shoe (see photos 1 and 2) at around 34 m depth. Coring operations were delayed to cement the loose shoe in place.

Photo 1: The drill bit is attached to drill rods inside the steel casing. The inner part of the casing shoe is fixed in place at the end.

Photo 2: When ready to drill, the outer part of the shoe is attached

and it is able to turn with the drill bit.

This morning we successfully drilled out the cement grout, through the shoe and into rock. Unfortunately a hydraulic hose blew during the first core run, so we are waiting to have it fixed.

We expect to resume today and then continue coring 24 hr/day until the hole is at its final depth. Our target depth is between 160 and 200 m, depending on geology beneath the fault, and finance.

We have had lots of public contact. I was live on three radio stations on Monday. Yesterday, we had newspaper photographers and TVNZ on site.

I expect there will be a story about us on Close Up tonight. Finally, we had a public meeting in the local village of Whataroa that was very well attended. There is clearly a high level of public interest in what we are doing.

Finally, I expect some preliminary results of our geological logs to be ready to report on soon.

Third Photo: A cement-bentonite grout is pumped through a Tremie tube to fix the loose casing shoe in place at the base of the hole.

Last photo shows a view of the drill site.

Core Impressive!

Just as I finished my previous update there was a sudden improvement in core recovery. We were at around 70 m depth and suddenly the cores looked very different – green, hard, and perfect; instead of falling to pieces with lots of washed out clay and residual fragments. Had we just crossed the main fault?

Egor, the driller, and some of the DFDP-1A team happily show off a perfect core of the principal slip surface of the Alpine Fault.

After we cleaned those cores and carefully logged them, we came to realise that we had not crossed the Alpine Fault, we were just getting the most beautiful cores. Maybe it was just the effect of increased confining pressure that resulted in better cohesion and recovery of the cataclasite. Our lab work, wireline logs and the DFDP-1B results should help to provide a scientific explanation for this change.

After 20 m of near perfect core recovery, we definitely did hit the principal slip surface. We were at 90 m depth. It was obviously the fault because there was Quaternary gravel lying beneath the thick sequence of cataclasite. The driller, “Egor”, was more excited than anyone – a perfect core across the critical zone. Fantastic!

The precious fault gouge core is now in a very secure box in a secret location and is marked “Beware of Virginia”. After reaching a depth of 96 m, we were having difficulty advancing further and for various reasons we decided to halt

drilling and then withdraw the drill string.

A perfect core of the active slip surface of the Alpine Fault from the DFDP-1A drill site, Gaunt Creek.

The project so far has been frantic, but efficient and very successful. There we were holding the first core of the Alpine Fault on Sunday evening and we were not even scheduled to start coring until the following Monday morning. We were at least 3 days ahead of schedule and it dawned on me, as project manager, that none of the other contractors were even here yet and some instruments had not even arrived from the USA. Elation started to turn into a mild panic. There was a lot of work to do and plans to change.

The revised plan and workforce has now started to fall into place, and the TV, radio, and newspapers have been on the phone all day. We reached a consensus decision today on where to site the DFDP-1B borehole and the percussion air-hammer crew has now showed up to start installing a surface casing at the new site. The current intention is to knock a 6” casing down to nearly 70 m, to try and avoid the difficult materials found in shallow parts of the DFDP-1A borehole. We are hoping to reach a total depth of around 200 m in our next borehole, if we hit solid rock beneath the fault. We start tomorrow morning.

Dave Prior decides to dig 200 m at the agreed site of DFDP-1B.