Introduction:Optical fibres have delivered a breakthrough for BP in the continuous monitoring of well performance in real time. Terry Knott learns more.
Breaking records and claiming firsts is not unfamiliar territory for BP's Wytch Farm oilfield in the south of England.
Over the past six years the operator has steadily pushed out the limits for extended reach drilling (ERD) from shore in the drive to access additional accumulations of oil lying under the environmentally sensitive preserve of Poole Bay in Dorset.
The latest ERD well to jump to the top of the league table in a string of record breakers, M16, is 11,278m long with a horizontal departure of 10,728m, drilled into the Sherwood reservoir last year to become the world's longest oil production well.
But now a different sort of breakthrough has stolen centre stage at Western Europe's largest onshore oilfield, adding a new dimension to Wytch Farm's complex wells by providing continuous surveillance of downhole production activity on a real time basis.
The innovation comes in the form of an optical fibre distributed temperature sensing (DTS) system which accurately measures temperature at one metre intervals over the reservoir section of a well, building up a picture of production behaviour and revealing downhole events as they occur.
Two wells have been monitored over a two-year period, clearly demonstrating the system's ability to yield information on the performance of each producing zone in the well in real time.
Identification of occurrences such as cross-flow between zones and behind-casing flow during shut-in, plus water encroachment, have also been made using DTS.
'We believe this is the oil industry's first permanently installed distributed temperature measurement system operating in real time,' says Nigel Leggett, technical director of fibre optics specialist Sensa, developer of the system.
'Normally, detailed information on the way a well is performing is gathered by production logging, using coiled tubing for extended reach wells.
But this is done only at infrequent intervals - maybe every two or three years in some cases - which provides only a snapshot in time, and does not enable you to spot events in the well as they occur.
Deploying optical fibres is not only more cost effective, but data is fed back constantly to help understand and improve well operations routinely.'
Using optical fibres as sensing devices is not new - development stretches back over thirty years with origins in the defence and aerospace industries.
But in the last few years interest has grown in their suitability for downhole measurements, given their advantages when compared to conventional electronic sensing devices, including reliability at high temperatures, greater sensitivity and intrinsically safe operation.
Sensa's experience with downhole optical temperature sensors began four years ago when the company - then known as Sensor Highway - began installing temperature profiling systems in steamflood wells in heavy crude oilfields in Canada and California, operating successfully at up to 280¼C.
Then in 1997, working with the company, BP initiated an optical sensors trial in one of its relatively simple horizontal side-track well completions at Wytch Farm - K-7 - as a forerunner for more complex ERD wells being planned at the time.
As in the majority of Wytch Farm's wells, K-7 is produced using an electric submersible pump (ESP), in this case located some 2000m into the well.
The trial was set up to determine the ESP's performance, particularly during conditions of startup, shut-in and low flow, when ESPs can be prone to overheating and potential failure, leading to costly workovers.
The well was also equipped with conventional downhole electronic temperature and pressure gauges, set some 50m above the pump in the wellbore, this separation being required to avoid signal distortion caused by electrical noise generated by the ESP variable speed drive.
The optical fibre system was installed in a 1/4in diameter conduit running from the base of the ESP to the wellhead, enabling the temperature profile to be determined over the entire length of the pump installation, rather than at only the temperature gauge location.
The data gathered established the relationship between motor, pump, inlet and outlet temperatures, and proved that conditions could exist where the readings at the conventional gauge could be misleading, permitting the pump to circulate while shut-in with subsequent overheating.
Encouraged by the results, BP opted for permanent DTS systems to be installed in two new ERD wells at Wytch Farm with the aim to measure temperature not for its own sake as in the K-7 trial, but to interpret this to provide continuous production data across the zones of the reservoir.
The first of the new wells, M-12 completed in 1998, is a complex dual purpose well, serving as both water injector and oil producer.
The well has a measured depth of over 6000m with a horizontal producing interval around 1000m long.
Oil is produced through a perforated 7in casing into a 41/2in annulus over the reservoir, which then crosses over above the reservoir interval through a cross-over sub into the 41/2in production tubing, to be pumped to the surface by a shrouded ESP (see diagram right).
Produced water is injected down the annulus, crosses over into the production tubing above the reservoir interval, and proceeds into the water zone below.
'M-12 is an ambitious well with a complex wellpath,' Leggett points out. 'Its configuration excludes the use of conventional production logging tools due to the shrouded ESP and the cross-over sub, hence DTS really comes into its own in these circumstances.'
The optical fibre was run with the completion through a 1/4in diameter conduit - a standard downhole control line - attached to the outside of the production tubing.
Two conduits were installed to form a U-shaped tube extending to the lower part of the well, below the pump.
To achieve this the conduits had to pass through the upper packer, for which Baker Oil Tools, suppliers of the completion equipment, developed a hydraulic wet connect/disconnect system with a smooth bore which allows the conduit to be installed during a single trip completion, and to be parted and reconnected if the ESP has to be recovered for maintenance.
Some 10,500m of optical fibre was installed in the conduit loop using a fluid drag technique.
The method used is an enhancement developed by Sensa under licence of the original British Telecommunications invention of 'blowing' an optical fibre along a line around 1000m long using compressed air.
For very long runs, as required at Wytch Farm, Sensa uses high pressure water at 8000psi to 'float' the fibre through the conduit at around 40m per minute.
Several attempts were made to install the fibre without success until it was discovered that the conduit had been contaminated earlier with hydraulic fluid, causing the fibre to stick to the conduit wall.
Flushing the conduit with isopropanol cleaned the bore, enabling the fibre to be installed without further difficulty, says Sensa.
Achieving the hardware installation was only one of the challenges in the project, as George Brown, then with BP and now Sensa's manager for interpretation development, explains.
'There was an element of risk involved as this had not been tried before and no-one knew if we would get meaningful results.
But we took the view that while temperature may not give you absolutely 100 percent visibility of how much production is flowing from which zone, in the absence of any other continuous technique, DTS has gone a long way to achieving the ideal.'
In the M-12 DTS system, a full log of temperature profile against depth takes about half an hour to build up through laser pulsing, plus a further quarter of an hour to analyse the 10,000 or so data points - speeding up the pulse rate can reduce the time to do this, but the 0.1¼C temperature resolution begins to fall off.
The signals are processed in a 'black box' at Wytch Farm, but are downloaded over a phone line by Sensa to its offices in Andover, Hampshire, some 70km away.
Arguably this is not truly 'real time' monitoring, but as temperature changes only very slowly in a well over periods of days and weeks, the company claims it is effectively real time, and certainly far more rapid than any other logging technique available.
In addition to observing basic changes in temperature profile by simple visual comparison of graphs, the temperature data are also fed into a proprietary thermal well bore simulator known as Wellcat-Prod, developed by Halliburton's Landmark business unit.
'Wellcat was developed to help obtain accurate temperature profiles for wells in the planning stage in order to design casing and completions more efficiently,' says Brown.
'It solves the energy equation for the wellbore and formation, and the equations of mass, momentum and energy for each flow stream in a multi-stream model of steady state or transient flow in horizontal wells.
This allows temperature profiles to be determined on estimates of three phases of flow for a multi-zone undulating horizontal well. It is regarded as one of the most sophisticated thermal modelling tools in the industry.'
Once a thermal model for a well is established, temperature profiles generated by DTS can be compared with this to identify divergence from the norm, for example temperature change caused by water or gas breakthrough. Conversely, the model can predict the temperature likely to arise from such events, enabling certain temperatures to be set as preconditions, or alarms. But as Brown observes: 'Of course in the Wytch Farm wells we didn't know exactly what we were looking for or what to expect.'
M-12 normally produces at 9500b/d from nine zones and also injects 16,000b/d of produced water.
Temperature profiles for the well were established using DTS with the well flowing, both with and without water injection, and during shut-in, when the temperature begins to return to the natural geothermal gradient of the surrounding sandstone in the reservoir.
When producing only, the increase in temperature caused by the ESP running is readily observed at 3700m, while at 700m, cooling of the oil takes place when gas is liberated as the pressure drops below the bubble point.
If water injection (70¼C) is employed during oil production, the effects of the ESP are seen to diminish, while the effect of gas breakout is masked.
Initial water cut from the well was around 10 percent, which chemical analysis identified to be water emanating from the reservoir.
The question was asked whether this water was entering the well from a known water finger at 4200m which was thought to have been successfully cemented-off during completion, or had water broken through in another producing interval?
Early in 1999, water injection was turned off for three months while steady state temperature profiles were acquired for production only, eliminating the counter-flow heat transfer effect of injection.
The temperature profile showed an increase coinciding with the cemented-off water finger, suggesting this was continuing to flow behind the casing and make its way to the producing zones.
The thermal well bore simulator predicts that 2500b/d of production is required from the water finger region to fit the recorded temperature data, with the remaining production, some 7000b/d, arising from the lowest producing zones.
But a 24-hour shut-in of the well showed the temperature change in the region of the water finger to be less than that predicted by the model, suggesting that the water is cross-flowing from one producing zone to another in the wellbore, changing the perception of the reservoir structure and enabling BP to plan a cementing repair when the well is next worked over.
In 1999 another new ERD well, M-17, was completed at Wytch Farm with a Sensa DTS system installed. This is a more conventional completion with optical fibre conduit attached to a 2.675in diameter tubing stinger hung below the production tubing and ESP, across the reservoir.
Soon after startup in October last year, an initial water cut of around 10 percent of the 5000b/d production was analysed as being injected seawater.
The DTS system revealed that in the first two weeks the temperature dropped to well below the geothermal gradient, suggesting cold seawater was entering the toe of the well from a seawater injection well, 200m away.
The injector had been in operation for some time but was previously believed to be separated from M-17 by a fault.
Turning the injector off and on was used to show the connectivity between the two, following predicted temperature changes and time of response.
DTS has showed up several useful and sometimes unexpected characteristics of M-17, including the source of produced water as it increased from 25 percent to 35 percent earlier this year.
In May, when the well was shut-in, it also became clear that the well cross-flows from the top perforations to the middle set of perforations, an unexpected effect.
The lower half of the well can also be seen to be warming back towards the geothermal gradient with time, as expected.
Nigel Leggett summarises the impact of the work to date.
'The DTS systems at Wytch Farm have demonstrated that while continuous temperature monitoring of a reservoir may not in itself be able to resolve every production problem, when used with other reservoir data and a comprehensive thermal well model, it is possible to develop a significant understanding of reservoir performance without the need for expensive and infrequently run production logs.
Early diagnostics with DTS mean you can spot a problem at the outset rather than find out too late that you have been losing production for months or years.'
He estimates that coiled tubing production logs at Wytch Farm can cost up to £500,000 a time. By comparison, a Sensa DTS installation for a long ERD well comes in at around £125,000, giving continuous data and no loss of production during logging.
The 'black box' signal processor - made by York Sensors in the UK, now owned by Sensa - is capable of servicing several optical fibre loops, hence multi-well systems, typically the closely grouped wells found offshore, would see a cost benefit. A four-well system, for example, would cost around £190,000, says Leggett.
As recognition for the breakthrough in downhole real time reservoir surveillance achieved using DTS, last month BP awarded its Wytch Farm team one of the company's annual technology innovation awards amidst competition from the operator's worldwide assets.
Sensa is already expanding the applications of DTS systems. In April this year the first DTS system to monitor gas lift valves came into operation in Shell Expro's Tern field in the North Sea, providing real time surveillance of three valves in the TA-27 wellbore at intervals down to 2825m.
Coming soon will be the second DTS system to go into BHP's Douglas field in Morecambe Bay, UK, which will see the optical fibre installed within an oil-filled hydraulic control line being used to activate zone control sliding sleeves in an intelligent well completion.
From farther afield, Sensa has also used its system to detect a downhole pump failure in Venezuela, through remote downloading to the UK of information from the signal processor at the wellhead.
Looking ahead, Leggett sees optical fibre DTS to be ideally suited to deepwater wells which are remote from host platforms, applications which will require the reach of optical fibre measurement to be doubled from the current 30km to 60km.
Optical distributed temperature sensing:
Temperature can be measured every metre along an optical fibre by sending pulses of laser light at a fixed wavelength down the fibre.
At every point in the fibre, light is back-scattered and returns to the source. Knowing the speed of light and the moment of arrival of the return signal, enable its point of origin along the fibre to be determined.
Temperature stimulates the energy levels of the silica molecules in the optical fibre. The back-scattered light contains upshifted and downshifted wavebands - the Stokes Raman and Anti-Stokes Raman portions of the back-scattered spectrum - which can be analysed to determine the temperature at origin.
In this way the temperature of each of the responding points in the fibre can be calculated, providing a complete temperature profile along the fibre's length.
(Another wavelength known as the Brillouin band, not used in the calculation, is also contained in the back-scattered light, which is a function of the temperature and strain on the fibre).
The 1064 nanometre wavelength laser light, or incident Rayleigh light in the spectrum diagram, is pulsed from both ends of the 250 micron diameter multi-core optical fibre every 8 nanoseconds. The returning light signals are gathered in the gaps between pulses. By sending light from each end, better data resolution is attained than in single-ended mode, giving temperature measurement accuracy of 0.1،C.
by Terry Knott
© 2001 Mena Report (www.menareport.com)
