The Need for an In-Line Oil-In-Water Monitor
Shell U.K. Exploration and Production, Aberdeen
In-line oil-in-water monitors have long been wanted for monitoring overboard discharges from oil and gas producing facilities, or for controlling water handling systems. In general their performance has fallen far below operational requirements. New techniques and steadily increasing demands for better measurement capabilities mean that systems are becoming available with the potential to give good performance. A review is given of oil-in-water measurement in Shell, application areas, and the difficulties in implementing the technology.
As part of the production of oil and gas, large quantities of water are also produced. Whether on land or at sea, it is generally accepted that produced water should be processed sufficiently that the quantities and concentrations of potentially harmful components are reduced to levels that are known or are deemed to be harmless to the environment in which the water is discharged. In the North Sea, the current protocol limits the content of discharged water to 40 parts per million (ppm) averaged over a calendar month. Over the years the relative quantity of water has increased as oil fields have watered out. North Sea water production is approaching one hundred million tonnes per annum, and consequently the current legislation permits some four thousand tonnes of oil per annum to be discharged from oil production facilities. This situation is replicated in most offshore facilities world wide. It is generally accepted that discharge of too much oil represents a global threat to sensitive marine environments, but there is not a clear consensus on what is "too much". Around the world, environmental and legislative bodies are trying to establish meaningful limits.
The situation is further complicated by the difficulty of measuring oil-in-water to part-per-million levels on continuously operating facilities in a hostile environment, where it is considered essential to keep operational costs, and hence manning levels, to a minimum.
What is "oil" in this context? Is it dissolved or dispersed? Are there particles coated with oil present? The current methods for detection and measurement of oil-in-water, namely chemical analysis of intermittent samples, or monitoring bypass sample lines by a variety of optical techniques such as fluorescence, absorption or scattering, respond in significantly different ways. Tests have shown that no single instrument gives a satisfactory solution for dealing with water disposal from current facilities.
Future facilities in the North Sea may require disposal of water from unmanned facilities, subsea facilities or even downhole. Oil-in-water measurement systems will have to operate at high temperature and pressure, and for other reasons than pollution monitoring. Active control of de-oiling systems is a clear example where there is no suitable instrumentation at present.
In this paper I attempt to set oil-in-water measurement in the general context of the measurements required for oil and gas production systems. I give a review of oil-in-water measurement as I have seen it over the last twenty years. I then discuss the needs for-oil-in water measurement, the variety of applications and the difficulties of implementing satisfactory solutions. I conclude by attempting to give an outlook for oil-in-water measurement.
2. Oil-In-Water as a Multiphase Fluid
Oil-in-water measurement is but one of the many measurements required in oil and gas production. Traditionally these measurements have been treated separately, and mostly without regard to the interactions with other parts of the production process. The development of multiphase technologies to transport and meter unseparated hydrocarbon streams allows and indeed forces one to take a more integrated view of the whole production process.
Figure 1: Multiphase Composition Triangle
The 'Multiphase Composition Triangle' shown in Figure 1 can be used to indicate conditions under which any measurement in the oil and gas production processes is made. We find single phase oil, water and gas at the vertices of the triangle; two phase fluids, oil/gas, water/gas and oil/water along the sides of the triangle; and the vast range of three phase fluids occupy the interior of the triangle. I have also shown a transition region, where the liquid part of the multiphase mixture may be either water-in-oil or oil-in-water, making measurements difficult for instruments using electrical properties of the fluid.
It is now easy to see that any oil and gas production process takes a particular multiphase mixture and performs a sufficiently good separation into marketable oil and gas streams and a waste water stream. How well these streams approximate to single phase flow depends on the process and how it reflects the specification of the oil and gas in the supply contracts and the environmental constraints for the waste water stream.
If we now focus on the waste water stream, we can immediately see that oil in water monitors may be required to control the water handling process as well as monitor the quality of the final discharge stream. Thus the Multiphase Composition Triangle is also useful in indicating where we have inadequate measurements and where we should direct our attention in developing new instruments.
3. Oil-In-Water Measurement Over the Last 20 Years
I have been associated directly and indirectly with oil-in-water measurement since the late 1970s and I thought it was worth giving my view of how oil-in-water measurement has changed over that time. This is not an exhaustive review of oil-in-water measurement, but simply a personal account of how and where I have been involved. Others may then be able to set that experience alongside their own, and hopefully may be able to take a better approach, or at least avoid repeating unsuccessful approaches.
In the late 1970s it was recognised that continuous monitoring of oil in water was desirable, but colleagues working with fluorescence based instruments could not get acceptable performance. Most of Shell Expro's offshore platforms had concrete storage bases, where most of the oil and water separation took place. With long residence times it was not difficult to achieve good water quality, and twice daily manual determination of oil content was adequate.
I spent most of the 1980's in the Production Measurements group at KSEPL, the then Shell E&P laboratory in Holland. We worked extensively on oil-in-water measurement for applications throughout the Shell Group, both onshore and offshore. We examined a wide range of equipment, for example automated chemical extraction and IR absorption systems, an Attenuated Total Reflection (ATR) system, various optical scattering systems, and particle counting systems. Of these, the automated chemical extraction system and the scattering systems worked best, but none of the systems was suitable for long-term unattended low-maintenance operation.
Interest in oil-in-water measurement waxes and wanes; new legislation in one part of the world results in demands for better equipment; lower oil prices tends to lead to delays in development or implementation. Without the stimulus of specific application needs and constraints it is difficult to provide practical equipment for field use.
Towards the end of the 1980s Expro wanted to develop the Eider platform as an unmanned satellite of North Cormorant, but were required to monitor continuously the oil content of the overboard water discharge, as twice-daily manual measurements were clearly impractical. A fluorescence based system operating on a bypass stream was considered best bet, but after very extensive evaluation it became clear that although it might make a reasonable thermometer, it could not even approach the target specification for oil-in-water measurement. We considered ways of automating sample collection as a possible approach to Eider's needs, but when it became clear that it was impractical to operate Eider as a completely unmanned facility the immediate need for a continuous monitor faded.
In 1990 the OWTC (Orkney Water Test Centre) carried out extensive tests of commercial oil in water monitors. The best performing systems were based on scattering, but after a detailed evaluation of the results KSEPL concluded that they could only be used reliably on unattended facilities as a coarse alarm at about 300 ppm. Interest at KSEPL then focused on improving scattering and IR transmission/absorption systems to the extent that they would be suitable for Shell requirements. A commercially available instrument has come out of that work, but it utilises a bypass sampling loop and cannot perform well at high temperatures.
In 1990 I was back in Expro and became aware of work being done by Heriot Watt University to develop photoacoustic instrumentation. In this approach pulsed focused infra red light is directed at a substance. The illuminated substance absorbs the radiation, heats up, expands and generates a pressure wave which is detected as an ultrasonic pulse. Water has a low photoacoustic response whereas oil has a large photoacoustic response. The immediately attractive features of this approach were that it would be possible to get a reasonably straightforward measurement from a sensor mounted directly in the discharge line, eliminating hard-to-maintain sampling systems. In addition, the response of the system should be sufficiently different to dissolved and dispersed oil to allow them to be satisfactorily discriminated.
Development of the photoacoustic system has proceeded to the point where Expro is ready to evaluate the system in the field. In 1996 there was pressure on several of our facilities to improve the quality of their overboard discharges. My advice was that the scattering systems that came out best in the 1990 OWTC tests and the fluorescence based system that performed well in subsequent OWTC tests would prove to be too difficult to keep working to specification, and that we should try to accelerate the development of the photoacoustic system. While the advice on the scattering and fluorescence systems has proved correct, the accelerated development of the photoacoustic system has not been without problems. It is fair to say that the measurement principle has been thoroughly verified - in the laboratory from 3 ppm to over 2% oil-in-water; at the OWTC in the environmental range of concentrations; and in simulated field conditions up to several thousand ppm. Nonetheless, there is not yet an ex-certifiable sensor with sufficient sensitivity that can be deployed offshore, nor have two sensors been made with closely similar performance. There is however good reason to believe that these difficulties can be resolved.
4. The Needs
Why do we really need oil-in-water monitors? The answer from a responsible oil and gas operator and from a responsible government is that too much oil released into sensitive marine environments will result in significant, even irreparable damage. But there is a major difficulty in knowing what is "too much", or indeed which components of the oil cause most damage.
How much do we really want oil-in-water monitors? If we have steady production and good separation capability, then manual measurements twice a day are easily good enough to keep track of things. But if, as has happened, the freons used in the manual extraction measurement are banned, and the replacements are more toxic, what then? If production is not steady and it has become necessary to have an operator take many more manual samples to get a better idea of what is happening in the process, then there is much more interest in a continuous monitor.
How much hassle are we prepared to accept from a continuous monitor? The answer to that one is fairly easy - not a lot. If a continuous oil-in-water monitor cannot give reasonably accurate, highly repeatable measurements of what is deemed to be oil (dispersed and/or dissolved) and at the same time be highly immune to contamination, suspended particles gas bubbles etc, we are unlikely to be interested in installing it.
As someone who has tried to stimulate development of continuous oil in water monitoring systems through several cycles of interest and lack of interest, the main requirements have emerged as follows.
- The measurement should be based on sound physical principles.
- It should be possible to make the measurement directly in the discharge, as we have had most problems with the sampling parts of oil-in-water measurement systems.
- Any surfaces involved in the measurement process should not be subject to contamination, or the measurement should not be affected by such contamination.
- It should be possible to operate at high temperatures. As our fields produce more water, the discharge water temperature has risen, and is about 100°C in one case.
- It should in principle be able to work at pressures above atmospheric.
- For environmental measurements, the system should be able to measure in the range 0 - 100 ppm with a relative accuracy of 10% and a resolution of 4 ppm. It should be able to discriminate between dissolved and dispersed oil.
- The technique should be capable of further development. Too many of the oil-in-water systems commercially available have already reached their technical limits.
I believe that a system that can comply with the above should be compatible with any reasonable standard that is likely to be adopted for oil-in-water measurement, and may even be a de facto standard itself. I further believe that in the photoacoustic approach that there is at least one approach with a good chance of satisfying the above criteria.
5. The Applications
In line with what I have said above, the safest approach, and I think the best one, is to treat each application on its own merits. The simplest applications are for fields where the oil comes from one reservoir, temperatures are modest, and the produced water has no peculiar properties. Several on-line systems may give reasonable performance in these circumstances. Let us complicate matters a little, and introduce a second reservoir. The oils are different. Scattering, fluorescence and photoacoustic systems all respond differently to different oils. Indeed, how well does the standard extraction method cope with different oils? How does one calibrate the system to read accurately when only one of the reservoirs is producing? The produced waters are different. Does scale form, resulting in particles (or worse) that affect the measurement? What does one do on a facility where there are multiple tie backs, when one can expect that there will be significant differences between the fluids?
Let us consider the produced water from gas fields and gas condensate fields, where there tend to be higher concentrations of aromatic compounds dissolved in the water. Many on-line systems do not respond to dissolved oil. The photoacoustic system responds strongly to the anthracenes present in crude oil. These anthracenes are absent in condensates, so the photoacoustic system will have to look for other components if it is to be used successfully for condensate in water.
Let us complicate things still further. We want to operate subsea or even downhole, and do all the above remotely at high temperature and pressure.
We would like to be able to control water handling processes better, for example hydrocyclones or tank drainage, so we must have oil-in-water monitors that can operate up to say, 5000 ppm oil-in-water reliably. A notoriously difficult problem is to measure the few percent of oil in wells at the ends of their lives. A direct measurement of oil-in-water up to say 5% would be valuable for optimising production from old fields. I would argue that the oil industry should be far keener to develop instrumentation for these applications as these can really save money. Oil-in-water monitors at the discharge point can only tell you to shut down your facilities if you exceed the permitted limits: oil-in-water monitors at critical points in the process can prevent the excess discharges happening.
I have not attempted to give a classification of types of applications, but have rather tried to show where we can use high performance in-line oil in water monitors to improve our production operations. I believe that it is now possible to develop such equipment.
6. Implementation - The Difficulties
It has been my experience that implementing new measurement techniques successfully in the oil and gas industry is in practice fraught with difficulty. Let us assume that the early hurdles of turning a good idea into a working prototype instrument have been successfully overcome. Several years will have passed, and perhaps some £250,000 in total will have been invested. Let us further assume that there has been at least active interest by an oil company and an equipment manufacturer so that the prototype equipment is compatible with oilfield installation requirements. Nevertheless, a field evaluation of a new instrument will probably cost more than the whole development to date, and may take over a year to organise.
In my experience if a field evaluation is to give useful data the, the staff on the installation on which the evaluation is to be carried out must be able to see direct benefit to them if the evaluation is successful. In these days of minimal manning, if there is no direct benefit to the people who have to support the evaluation, they are unlikely to devote adequate time, especially when things go wrong. A further consequence of minimal manning is that for field evaluations today, it is almost essential to provide means of getting the data onshore so that the evaluation can be monitored remotely.
After satisfactory field evaluation, one is then faced by the difficulties of turning working prototypes into fully commercial instruments. In the case of oil-in-water monitors, one really requires considerable feed back from instruments operating in the field to confirm that the instrument performs correctly. Conditions in the field are quite different to those one can simulate in a test loop. It is often stated that operating platforms are not places on which to conduct R&D projects. Nevertheless, if one wants to gain the benefits that better measurements can undoubtedly bring, one cannot exclude operating facilities from the R&D process.
In saying the above, I am not making excuses for why we do not yet have satisfactory oil-in-water measurement systems, but I am trying to point out that many groups of people, academics, industrial researchers, government departments for environmental matters and also those for stimulating development, manufacturers, engineers and operators in oil companies and their contractors must somehow form an extended team for about ten years if a successful conclusion is to be reached. In my experience, such teams are not put together. They simply happen because those who decide to be members recognise that they need to be involved. I continue to be surprised at how effective such extended teams can be at getting things done while they can work together. It is at the implementation stage, when large amounts of money must be spent, that these extended teams are most likely to break up. If there is no longer a clear need expressed by a keen prospective user and that user's active involvement, there is virtually no incentive to continue. Development languishes until another potential user expresses enthusiasm, and all one can do is hope that earlier work has not been wasted.
7. The Outlook
Oil and gas producing companies have wanted cost effective continuous oil-in-water monitors for many years in order to have good knowledge of the quality of their produced water discharges. The systems available to date have not shown good performance, and require intensive maintenance. Most systems are operating close to their technical limits so it is very unlikely that they can be improved sufficiently even if large sums of money were spent in further development.
In recent years the photoacoustic approach appears not only to have the potential for measuring accurately in the environmental range, say 0 - 100 ppm, but also up to several percent, so that that approach could be used for control of compact water treatment systems, or even for monitoring the oil content of high watercut oil streams, potentially allowing better optimisation of production facilities at the end of their lives. This particular technology is in its infancy, but, in my opinion, can already outperform most other oil-in-water systems from a measurement point of view. However, there is a long way to go before there will be assured operational performance.
In the North Sea of the relatively near future there will be many different types of facilities; the early large fixed platforms, minimum topsides facilities, floating production systems, subsea production systems and even complete downhole separation facilities. These will require a wide range of oil-in-water measuring devices to cope with the very different installation requirements and the different physical and chemical properties of the produced water. For example, subsea and downhole systems will probably have to contend with high pressures and temperatures.
I think that there is a large potential market for cost effective oil-in-water measurement systems, with several on each facility for monitoring and control applications. This worldwide market is far bigger than can be serviced by one manufacturer, or indeed one measurement technique. However, for this market to develop there needs to be consistent confirmation from users and society as a whole that increased use of better monitors brings worthwhile benefits.
Comparing oil-in-water measurement Varying government regulations and measurement methods call for standardization Colin C. Tyrie, and Dan D. Caudle - There are several instrumental methods for measuring oil in produced water. None of them measure all the organic compounds in the water. Comparing what the commercially available methods actually measure will illustrate the problem of interpreting oil in water (OIW) analysis. From worldoil.