Abstract
Three phase gravity separators, degassers, hydrocyclones and coalescers are commonly used in
the oil and gas industry as primary and secondary means of separating well-head fluids, which may
appear as a mixture of gas/oil/water. Although the design and principles of operation behind these unit
operations are well known; assessing their performance as part of an integrated system to predict oil in
water, water in oil, and perhaps liquid in gas flowing out of them is complex due to the variability of
the inlet flow pattern, the fluid characteristics entering the production facility, and due to the difficulty
in predicting the droplet size distribution entering the system and the droplet size distribution inside the
separators. A system model capable of designing and assessing the performance of the production
separation system of an offshore platform regardless of the design configuration and arrangement of
the production equipment installed has been developed to overcome these complexities; no such
system model has been published before.
A design model for primary gas/oil/water gravity separators has been built using a
comprehensive procedure found in the literature, and the results have been compared to the installed
industrial separators in the Alba Field in Equatorial Guinea, property of Marathon Oil Corporation.
Results obtained showed that the design procedure in the open literature may be a little bit more
conservative than that used for primary separators in production fields, mainly when setting limits for
the holdup and surge times in the oil and water compartments of separators. However, it can be
concluded that the procedure found in the open literature does correspond to that of primary separators
in production fields. Assessment of the design of the Alba Field primary separators showed that the
field separators will allow water in the oil stream leaving the separators due to insufficient oil
residence time in the gravity separation area, while insufficient surge time in the oil bucket may result
in oil spill over the oil bucket to the water side and out of the separator with the water stream.
Inadequate holdup times in both the oil bucket and water compartment could also result in gas leaving
the separator with both the oil and water streams. Evidence for all these potential problems has been
observed in the downstream production facilities.
A performance model has also been built and used to assess the performance of the Alba Field
separators. The results obtained have been published (Abia-Biteo Belope & Thorpe 2007). Results
from the field when cross-plotted against the performance model produced a clear relationship between
oil contain in water and water contain in the oil streams leaving production separators. The
performance model results also produced a guideline on how the Alba Field separators should be
operated depending on the degree of separation required for the downstream processes. It is hoped that
the general method of plotting performance versus a dimensionless variable (θ / t) may be widely
applicable and may in time yield a universal curve for all separators. Finally, predictions of the model
for changes to key separation variables such as flow rate, temperature, pressure and increase in
dispersed phase droplet size are found to be consistent with results from the literature.
The effectiveness of the dissolved gas flotation technique for the treatment of oily waste water
from an oil production water system had been demonstrated in the study by Hanafy & Nabih (2007), in
which they showed that high separation efficiencies of about 80% could be obtained for the removal of
oil droplets. However, residence times in the order of 20 minutes were required, which are simply too
high for offshore production platforms due to limited space. An experimental study was conducted at
the University of Surrey, using part of a dissolved gas flotation rig previously used to investigate the
flotation of solid particles (Hague 2002) and the determination of the local rate of release of air in the
contact and flotation zones of a dissolved air flotation tank (Fontain 2003); to investigate the efficiency
of the dissolved gas flotation technique at residence times comparable to those of other gravity
separation equipment installed in offshore production platforms (3 – 6 minutes). The study consisted of
creating an emulsion of droplets of size less than 100 μm in a continuous water phase using a Sulzer
mixer, and then separating the emulsion using gas micro-bubbles of an average diameter of 50 μm with
standard deviation of 5.5 μm. Results of the study showed that the principle of dissolved gas flotation
could be successfully applied to offshore degassers to separate oil droplets of size less than 100μm,
which may have been missed by the primary separators in an offshore separation process, achieving
efficiencies as high as 80% at residence times of about 5.5 minutes. The results have also provided, for
the first time, plots showing the relationship between droplet size and separation efficiency in a
dissolved gas flotation system. Finally, it is hoped that the study conducted could also be used as a
platform for further studies that could prove that oil separation efficiencies comparable to those
achieved in hydrocyclones (> 90%) can be achieved using the dissolved gas flotation technique, if the
right gas bubbles size distribution to attach a known oil droplets size distribution are present in the
water stream, and if the required gas saturator pressure and gas saturated water flow rate are set.
The offshore separation system model built also incorporate the prediction of the droplet size
distribution throughout the production system, from the choke valve to the water stream to be
discharged to the sea or to be injected back to the reservoir. Results obtained when applying the
maximum stable droplet size distribution model of Van der Zande et al (1999b) to the Alba Field
system showed that very small oil and water droplets in the region of 10 μm are present in the
production flows, which may not be separated in the primary production separators, as they are much
smaller than the critical droplet size often designed for these separators, which is about 100 μm. Some
of these droplets may coalesce in the production pipeline between the choke valve and the primary
gravity separators. The model by Rozentsvaig & Pergushev (1981) was used to estimate the maximum
droplet diameter likely to leave the production pipeline as a result of coalescence due to turbulent
stirring during motion through the pipeline. Results obtained when applying the model of Rozentsvaig
& Pergushev (1981) to a surface production pipeline between the choke valve and the primary
production separator in the Alba Field showed that the small oil and water droplets generated across
the production choke valve will coalesce to sizes in the region of 60 μm and 30 μm respectively.
However, these droplets will still be smaller than the critical size for separation in the primary
production separator. If sufficient residence time is not allowed in the production separator for these
droplets to coalesce to a size equal or larger than the critical size, then they are likely to be carried out
to the produced water cleaning system where they may be separated, depending on the efficiency of
the produced water cleaning system. The current produced water system design in the Alba Field is not
efficient for droplets of such size. The introduction of dissolved gas floatation to the existing degassers
could provide efficient separation of such droplets.