The big three oilfield service providers believe the new generation of formation evaluation tools will answer the oil company's every prayer.
Rick von Flatern checks out their credentials.
As the oil and gas industry has moved farther from shore and into water depths thought impossible only a few short years ago, it has also entered a world of unprecedented drilling costs. Deepwater drillships and semis run at around $500,000 a day while on location.
The name of the game for cost-conscious operators then is to minimize days on site and especially those days spent in the flat sections of the drilling curve when time is moving but the bit is not.
One of those times when no hole is made, but which those who pay the bills dare not skimp on, is flow testing. Traditionally, these well tests are run for several weeks and in some cases, in the form of extended well tests, months.
Their cost is borne with relatively little grumbling as the only reliable means of getting the kind of accurate reservoir data used in every facet of field development.
'When an oil company finds a reservoir and wants to evaluate for investment it has to put a team together,' says Schlumberger section manager Andrew Kurkjian. 'And everyone needs to know pressure, flow, composition, what kinds of problems to expect. Everyone planning the development needs a formation fluid sample for design.'
But as well tests run to the tens of millions of dollars, operators have become decidedly less sanguine about them. Forward-looking operators like Shell in the Gulf of Mexico began searching for more economical ways to gain essential reservoir information as early as the mid-1980s.
But those early efforts, generally directed at improved formation testing tools, were abandoned. Drilling fluid contamination and uncontrolled drawdown volumes and pressures, combined with the inability to maintain fluids in reservoir condition as they were brought to the surface rendered the reservoir fluid samples woefully inadequate.
Now, at least three of the industry's leading service companies - Halliburton, Baker Hughes and Schlumberger - say they have solved those problems and are ready to bring to the market what are essentially the next generation of formation test tools.
Schlumberger's tool is called the MDT, Baker's the RCI and Halliburton's the RDT tool. They have the potential, say their creators, of garnering the same or better reservoir data as traditional flowing well tests in deep water at a tenth of the cost.
The first challenge to reservoir fluid sampling derives from contamination of reservoir samples with drilling fluid filtrate. While it is relatively easy to take a sample beyond the skin left by mud additives on the borehole wall, base fluids that enter the formation itself mix with produced fluids. The problem becomes especially sticky when, as is the case in all deepwater wells, synthetic oil based drilling mud is used and the base fluids are indistinguishable from reservoir hydrocarbons.
The three service companies have approached the problem in two ways. Halliburton has chosen to expand on its earlier perforate and sample system and to get uncontaminated systems through large volume sampling.
'We cannot guarantee any more than an MDT tool to get behind filtration,' says Halliburton Energy Services' Jay Burris from his offices in Dallas. 'The big difference is that we can have a sample chamber all the way to the surface. Most customers feel confident they are getting good samples if we can get 20 barrels.'
Schlumberger and Baker have taken another route using multiple, smaller sample bottles and an optically-based fluid analyzer, essentially a light spectrometer that analyzes the optical density of fluids flowing through the tool.
By passing a light wave of known frequency through the fluid its density can be identified by the amount of light absorbed. Once flow is initiated then, since all hydrocarbons absorb light at different rates, the contaminating base fluid can be discerned from reservoir fluid (disallowing the minute possibility they are exactly the same) as the amount of light reaching the sensor changes with time.
'What happens is that you start with all one kind of oil and end up with another,' explains Schlumberger's Kurkjian. 'We have a realtime sensor showing us the clean up time and that is why people are becoming comfortable with it.'
Though Baker uses the same optical density measurement approach to clean up, its engineers differ from Schlumberger's in one critical area - determining in realtime the degree of contamination as a percentage of fluid passing through the tool.
'The only way to quantify contamination in the sample is to have a sample of drilling mud as it was when it cut the formation and a sample of the hydrocarbon that is in the formation,' says Baker's deepwater business development manager Rod Larson.
'You need to have those two end points to determine percentage or it is based on modeling that is only as good as the assumptions you made in building the model.'
Schlumberger, it would seem, has faith in the math. Using realtime feeds from the spectrometer and extrapolating a curve of the apparent change in fluid composition over time, it says it can accurately tell operators the percent contamination of the fluid as it passes through their MDT.
'The OFA (Schlumberger's title for its light spectometer) is 10 years old and the product it got originally was oil, gas and water,' recalls Kurkjian. 'About four years ago people realized they could get a clean up but not until this year could we put a number to contamination in realtime while you were pumping out.'
According to Schlumberger RS&P product champion, Troy Fields, the same method makes it possible to determine with reasonable accuracy the time required to reach a desired contamination level.
'Using this method we can actually predict within your acceptable time limits when to sample,' he says. 'We have to be into the flow time a little and as you wait and get more sample through the tool, you get better prediction.'
How long it takes to reach an acceptable level of contamination before switching flow to the sample bottles is of course a function of the operator's required level and time/cost restriction. For most operators the level is somewhere around 5 percent contamination.
Drawdown and sandface control:
In traditional formation testing methods, the sample chamber is at atmospheric pressure. When the probe is set and the chamber opened, the fluids rush in to the much lower-pressured bottle.
The sudden differential at the formation, however, causes severe drawdown pressure and, in unconsolidated formations, it can even cause sandface which in turn can cause failure of the probe-formation seal essential for minimizing mud filtrate contamination.
Most importantly, for the purposes of analysis, the sudden pressure changes distort the characteristics of the sample so severely it is all but useless to labs trying to form an accurate characterization of the fluid as it exists in the formation.
To combat uncontrolled drawdown, Halliburton has equipped its tool with a downhole choke that can be manipulated from the surface through telemetry the company has developed in the course of other research projects.
The competition uses what amounts to a downhole manifold through which samples can be turned to the sample bottles when they are deemed sufficiently pure by whatever means.
In both the Schlumberger and Baker systems, produced fluids are pumped (as opposed to merely allowed to differentially flow) through the tool and into the wellbore fluids. As a result, the back-pressure on the well bore is equal to the mud hydrostatic.
The sampling bottles are connected to the bypass line at a point between formation and wellbore. When it is time to retrieve the sample, the bottle is opened and the bypass valve closed.
Schlumberger calls the process 'low shock sampling' having been once told by unconsolidated sand experts that the old method 'shocked' the formation.
'All low shock sampling means is the mud keeps the sample at well pressure,' explains Kurkjian. 'All you do is open the valve and the flow rate and pressure remain the same.'
Baker's RCI tool works with an essentially identical system of pumps and pistons that retract against wellbore hydrostatic when flow is turned to the sampling bottles.
'When we begin to pump the rate is controlled to keep above the bubble point and not cause excessive drawdown in the formation,' says Baker's Larson. 'The floating piston in the tank slides back and dispels the wellbore fluid behind it. When the piston hits the stops, we are able to overpressure the sample well above the formation pressure.'
Similarly, Schlumberger adds nitrogen charging to its sample bottles. They both do so to combat yet another problem associated with first generation formation sampling techniques - the change in pressures that occur in a closed system when a sample is cooled as it is being brought to the surface.
As a rule of thumb, pressure increases about 65psi for every 1°F rise. Since the opposite must be true, sample cooling of only 100° between formation and wellhead results in pressure drops of about 6500psi, a difference that easily causes the kind of fluid character changes that occur during overly aggressive drawdown. The added pressure essentially acts to shrink the volume of the sampler space to minimize pressure drop.
The final leg of the next-generation formation test tool is transportation. So much effort to retrieve a pristine sample is obviously all for naught if it is contaminated by a pressure or temperature breach between the wellhead and the lab.
Also, even in oil industry centers like Houston, the lab can be sufficiently busy to require a month before it opens it and discovers it has been compromised - a considerable setback to operators in terms of both time and money wasted.
Schlumberger has opted to tie its transport wagon to Oilphase, an Aberdeen-headquartered division with a long history of sample maintenance, handling and analysis. Baker testers say their company has opted to remain free to avail itself of any commercial transport experts for flexibility.
'To make a long story short, the MDT guy has a piece of equipment downhole that is real sophisticated,' says Schlumberger's Kurkjian, explaining the separation of duties at the surface.
'So he has his hands full and is not necessarily trained in PVT (pressure, volume, temperature) analysis but the Oilphase engineer is, so the two work as an integrated team. Also, if clients want a second validity check of the sample at the surface, they can do that and get a quick PVT analysis without waiting a month for the full laboratory analysis.'
Schlumberger also reheats the samples to downhole temperatures before transferring them to transport bottles, while Baker has avoided transfer pitfalls by eliminating transfer altogether.
'Although we are using pressure to compensate for shrinkage, the advantage of these tanks is the sample need not be transferred,' explains Baker's Larson. 'The risk of contamination from other transport tanks and the transfer method itself is eliminated'.
His colleague Mike Shammai adds: 'There is also a manual valve on each cylinder; when taken out of the carrier they are closed, so the sample will be safe for transportation.'
Replacing the flow test:
In the high stakes game of deepwater exploration and production, finding new hydrocarbon reservoirs is but the first of many, expensive steps on the way to new production. With the size of investment required to turn a discovery into a development in waters beyond the continental shelf, defining as precisely as possible the rate and type of production and ultimate recovery possible is critical.
As a result, operators find themselves faced with a conundrum. Since so much is riding on obtaining accurate reservoir information, they are most comfortable with the large of amount of data available from traditional extended well tests.
But with the costs for such operations now running to tens of millions of dollars there is almost as much of a gamble in running a test as simply drilling a second well.
The new formation testing tools offer an answer to the dilemma, according to HES's Jay Burris, particularly when used in conjunction with the measuring tools perfected in recent years.
'With the right type of logging run and the right type of seismic run operators are going to know 95 percent of the information they need to make development decisions about their reservoirs,' he believes. That would appear to be just the answer operators need.
By Rick von Flatern
© 2000 Mena Report (www.menareport.com)