Delivering the perfect formation fluid sample
  from: Asian Oil & Gas
  by: Rick von Flatern
  Thursday, August 24, 2006

With increasing water depths, sound decisions on how, when or whether to complete high-risk wells have become ever more critical to operators' economic health. In recent years, with their constant refinement, wireline formation pressure testing and sampling tools have become nearly indispensable in that decision process. Rick von Flatern spoke with service providers about how they plan to deliver the goods.

Perfect results - zero contamination, exact maintenance of in situ conditions - are rarely if ever achieved and so processing has long relied on corrections, correlations that depend on certain assumptions, all of which are potential error sources.

For most of downhole sampling tool history, these uncertainties were deemed acceptable as even flawed tests provided operators with critical knowledge early on about such things as fluid composition and viscosity.

But today, formation testing experts are determined to deliver more precise data as the cost of mistakes in remote locations, hostile environments and deep water continues to rise.

Traditionally, obtaining the cleanest sample possible was achieved by flowing the well through the sampling tool into extra bottles on the assumption the last bottle filled would be mostly formation fluid. This system was refined when tools were able to flow to the annulus until sensors on the tool determined contamination levels were at acceptable levels.

Over time that acceptable level was as high as 10% contamination, even more on stubborn wells. And though better than no sample or pressure test at all, margins of error can grow with contamination levels.

Too, as is the nature of any tool in open hole, the longer a tester remains on station the greater are its chances of becoming stuck and the higher the cost of the job in terms of rig time - no small consideration given the current state of dayrates.

Better with age
Halliburton uses its MRILab to provide a direct measurement of live oil flowing from the reservoir as it passes through its RDT (Reservoir Description Tool). It also provides, according to company literature, real-time, laboratory-quality fluid characteristics, such as gas-oil ratio and viscosity.

'The MRILab is maturing nicely,' says James Buchanan, Halliburton's product line manager open hole logging. 'It is a complex product and we needed an extended period of time to mature it. We are at the mature point this year. A significant portion of our testing budget has been directed over the past several years at refining it and improving the answer product the company sees.'

Halliburton runs the MRILab tool above the sampler. As the RDT pumps formation fluid out of formation, the flow path of the fluid is going across the MRILab sensor and with NMR techniques built on T1, T2, hydrogen index and viscosity that drives contamination algorithms, the operator can monitor contamination levels in the sample in real time. 'When we are satisfied we are cleaned up, we actuate the valving that permits the sample to be put into the sample chambers', says Buchanan. 'If you pump long enough you can get down to 7%, 5% contamination. It is the rare formation and mud cake condition that will let you achieve zero contamination with these pump out testers.'

Once the sample has been deemed to be sufficiently uncontaminated it is monitored to guard against change of state by keeping it at a pressure above the bubble point when sampling oils or the dew point for gas samples.

'We take great care to put the undisturbed sample in the bottle such that even when it is being brought to the surface and put on a helicopter and then sits on somebody's shelf for three months before it gets analyzed, its state will not change,' Buchanan says.

To transfer formation fluids from the formation to the sample bottle Halliburton has developed several variations on the traditional probe. Last year it finished development of a straddle packer technology to be run in conjunction with its standard dual probe.

'We added a couple new wrinkles to our straddle packer technology,' Buchanan says. 'As you might imagine when you inflate a packer in open hole and it sits for an hour or eight or 12 while pumping, one might get concerned when it is time to move on to the next sample. So one of the important and unique features of the tool is we have a "power close feature" so that when we are done testing, the packers deflate and go soft and then we actually pull them closed. It is definitely new and should translate to much lower risk of sticking the tool and extended life on the packer.'

The new tool also has two intake points between the packers as opposed to the traditional single. 'Multiple intake points allows us to do some interesting things,' Buchanan says. 'For instance, we can circulate between upper and lower points and that might be helpful in doing some very near well bore stimulation, removal of mud cake, for instance. We could maybe bring down a payload of surfactant and clean up what otherwise could be a nasty mud cake situation and when we are done we can go about our business of testing.'

The system could also be used to fill up the annular space between the packers, he says, to check for fluid segregation when the RDT's pumps are turned off. And since the standard probe is being run above or below the straddle packer, depending on formation conditions, the operator has the choice at each station of deploying either type probe.

The standard dual probe has also been redesigned to include an oval-shaped pad that encloses the two probes with a footprint of about ten inches that comes in contact with the formation.

'It is an oval donut-shaped packer with a channel between the two probes,' says Buchanan. 'It is in fact like a very small straddle packer. In particularly highly laminated environments, if you are not lucky, you can put a probe directly across an impermeable shale streak and unset to try again. But try moving a logging tool three inches. This level of precision is really tough to achieve. With this you get around that whole problem because you are across several dozen mm or cm thick layers and achieving what would otherwise call for a straddle packer operation.'

Customers value it, he says, for its reduced risk and deployment time compared to straddle packers. 'If you can get it done with the oval pad you just saved the customer three hours rig time, six hours. It adds up and really adds up if you are in a sticky hole. If the customer has asked us to go in with a straddle packer the standard RDT probes in most cases will also be going in with the oval pad on them. It is a very simple, elegant solution rather than the brute force solution of a straddle packer.'

Larger than its parts
Focused around its RCI (Reservoir Characterization Instrument) Baker Atlas has taken what might be described as a holistic approach to sampling and pressure testing and bundled it under a reservoir fluid characterization service called RESolution.

'RESolution is a service and the premise behind this service is that we are moving into a new era in formation testing and sampling,' says Baker Atlas product line manager, Michael Shammai. 'We have to serve the client with high quality samples and pressure tests. These are absolutely crucial pieces of data the client needs to have in order to evaluate the economics of the well and to evaluate the development and production of the well.'

Shammai says the goal is to define for the client the uncertainty of the data by making him understand exactly the procedures used and in so doing define the sample and pressure test quality that, in turn, defines the accuracy of the calculations the operator is doing on the data.

'The other aspect is that we have to get a representative enough sample because what (operators) do is send the samples to a lab and go through a procedure called a crude assay, essentially a mini-distillation column to break the oil into its components,' he says. 'That tells the client how much of a certain grade of gas or jet fuel they can get and that would determine the value of reserves.'

Such an early analysis can also tell operators whether their crude may 'poison' refineries with heavy metals such as mercury, selenium, barium, or H2S, all ingredients that can also affect reserves value.

'So our clients concern goes all the way to the refineries,' he says. 'After all, the purpose of them doing this is to make money and we have to understand how they make money and that is what RESolution is about. Until now we have had the information in blocks but this provides them the whole picture.'

Key to providing their clients the whole picture is, of course, good sample gathering. In the RESolution scenario, they begin by connecting logging while drilling data (when available) to the wireline job through pre-job modeling.

In pre-job modeling, RESolution uses TesTrak or other LWD data to type the sample and to predict such characteristics as porosity, permeability and zone thickness. If the LWD option is unavailable a set of correlations developed from an extensive crude oil database to which it has access through an agreement with GeoMark can be used to determine expected sample type according to the area being drilled and the source rock. Shammai says, the approach provides a fairly accurate indication of what to expect about 80% of the time.

The RCI has two packer module probe types - a traditional single packer module that uses a conventional probe that attaches to the reservoir and a straddle packer module. 'We model these two processes whose dynamics are very different,' Shammai says. 'The time taken to do these two processes is important because our customers often take numerous pressure tests. So if you can reduce time per test, reduce time down hole, you can reduce costs clients are paying for high price rig time.'

The operation includes what the company terms 'precision pressure testing' which takes into account the accuracy of the gage, depth control and a mathematical technique called FRA (Formation Rate Analysis) that evaluates the pressure test for Darcy flow.

'It is a very, very important aspect of RCI that allows our customers to determine the uncertainty of our pressure testing,' Shammai says. 'If you are dealing with Darcy flow you can say your pressure testing is very good or your mobility is very good. But if you do not have Darcy you have to look at the repeatability of the pressure testing. And see how that is repeating and if it stays within a window.'

The other aspect of a good pressure test, according to Shammai, is temperature stabilization to make certain the gage is unaffected by temperature changes from one depth to the next. 'We have a rigorous temperature stabilization procedure that in affect removes that variable ahead of time,' he says.

Baker Atlas calls its fluid capture operation 'Smart Sampling'. It begins with monitoring the sample as it moves from the formation through a conduit to the tanks, assuring it does not drop below its saturation pressure, a feat accomplished through control of the RCI pump rate.

From pre-job models and gradients gained through its pressure testing, RESolution can predict sample saturation pressure and viscosity through a process it calls PVTMOD.

'These two elements tells us if we can remove the sample from the reservoir in single phase,' says Shammai. 'That sample, under a controlled pump rate, is pumped into a tank that keeps it in single phase.'

The company provides two single phase tank types, Single Phase I, rated to 13,000psi and 250°F and Single-Phase II, rated to 25,000psi and 350°F. It retains the samples at single phase to the surface and beyond using nitrogen to overpressure the tanks.

Once at the surface a device called CDR (Continuous Data Recorder) is attached to the tanks to constantly measure and record pressure and temperature in the tanks from the time it is removed from the RCI in the field to the lab under the watchful eyes of Petrotech, a Norwegian company that is now Baker Atlas's official laboratory and who specialize in sample transfer.

To determine a fluid's purity level during sampling the RCI is deployed with an optical section called SampleView that measures optical density, refractive index and fluorescence.

'An important aspect of the design of the RCI is that it allows flexibility for different conditions downhole so you can assess the situation and react to it,' Shammai says. 'If you don't know what type of sample you are going to get downhole - black oil, volatile oil, condensate, retrograde gas - your tool has to have enough mechanical aptitude built into it to react to that situation to produce the highest bulk of sample for the client.'

RCI can mitigate against phase change during the sampling process by altering the pump rate down hole. 'You are evaluating pressure through FRA and PVTMod predicting sample type,' says Shammai. 'You are also monitoring the mobility change while pumping, optical density, fluorescence, refractive index and calculating gas-oil ratios and API gravity and it is all streaming up to a satellite connection or in the field in real time to the engineer. The engineer then makes a judge on what kind of sample he is getting and what should he see when he overpressures it in the tank. Changes can be done automatically right away.'

Lowering the bar
If Schlumberger's Tribor Rakela is to be believed, the levels of contamination operators will accept in their samples may soon be scaled down dramatically.

'The Quicksilver Probe is game changing technology,' the product champion says. 'It is making it possible to achieve a pure sample in a very limited time. And by pure sample I mean very low or even zero percent contamination.'

Such low contamination percentages, Rakela says, will enable downhole fluid analysis sufficient to provide real time basic PVT measurements that can have significant impact on reserves estimation, production rate, conductivity and flow assurance.

And while in many cases such purity is probably not critical, Rakela says some of the company's Gulf of Mexico clients have been very focused lately on having a less than 2% contamination level because of concerns over the uncertainties of viscosity measurements given them by the laboratory.

'Viscosity is very critical for flow assurance and to be able to tell if the well is going to be able to produce,' he says. 'In those cases we spent more than 18 hours trying to get a sample with the standard (Schlumberger's formation test) MDT tool and we could not get the levels better than 6%. So we went with this tool and after six hours we got 1.8% contamination.'

The main advantage of such pure samples, he says, is that in deepwater and other high-cost, high-risk jobs, the magnitude of the penalty for planning mistakes caused by contaminated samples can be on the order of millions of dollars.

The Quicksilver Probe differs from its MDT predecessor in the design of the probe itself and is driven by the fact filtrate will flow faster to the sampling point than will formation fluid because of vertical and horizontal permeability.

The solution then, says Rakela, was to split the flow into two paths with one sampling point isolated from a surrounding 'guard area' probe. The concentric probes are connected to discrete flowlines and pumps with full analysis capabilities on each. The result is filtrate moving along the wellbore wall is taken in at the 'guard' probe while formation fluids flow directly to the central probe.

'We are creating a cone that is going to be bringing the formation fluids into the sampling point while all the contamination flowing in from the sides is going into the guard area,' Rakela says. 'Essentially you have created an easier path for the contamination.'

In essence the tool uses the dreaded 'coning affect' turning what is normally a serious problem in water drive reservoirs to its advantage by using the tool's pumps to create intake velocities at the guard probe that are three to five times higher than those at the center take point.

Three years in development, the company reports it has done numerous jobs to date in India, Nigeria, Gulf of Mexico and the North Sea and says it was unable to get less than a 2% contamination level in only one instance and in more than 20% of the cases it achieved levels that, if not 0% contamination, were nonetheless too low to measure in the lab.

The service also has shown another advantage, Rakela says, through efficiency related to the time of operation. 'With this new probe we are going to be able to take significantly cleaner samples a lot faster than previously possible,' he says. 'Quicksilver Probe is as much as ten times faster and ten times cleaner, depending on the formation and fluids properties.'

And in the North Sea and Gulf of Mexico, Rakela reports, some operators have been able to reduce the number of samples they have taken as very low contamination levels allow them to scan fluids as they are pumped. They can then evaluate them for such information as GOR and basic composition and decide they do not need more samples.

Commercialization of the Quicksilver Probe was completed recently. AOG

Click here to register to receive your own copy of Asian Oil & Gas each month.