Boosting heat transfer on the Dalia FPSO

Total has used a combination of shellside and tubeside enhancement to reduce the size, costs and fouling potential of heat exchangers on the Dalia FPSO in the Gulf of Guinea. The enhancement technology saved around 269 t in wet weight and cut the plot area needed by 75 percent (78 m²).

Keeping down equipment size and weight while improving reliability is always critical offshore – and never more so than when operators are pushing the limits of current technology.

A prime example is the Dalia field in the Gulf of Guinea off the coast of Angola. Thought to contain up to a billion barrels of oil, Dalia extends over 230 km² at water depths of up to 1,500 m. The oil is viscous, sour, and hidden in complex geological structures under more than 700 m of unstable sediment.

Connected to the seabed by eight risers and 75 km of umbilicals is one of the world’s largest floating production, storage and offloading (FPSO) vessels. Measuring 300 m long, 60 m wide and 32 m high, the 416,000-tonne FPSO can accommodate up to 190 people as well as 2 million barrels of oil, with a processing capacity of 240,000 bbl/d. Despite this huge size, however, space and weight are at a premium on the FPSO, as with any other offshore installation.

Heat transfer challenges

The oil/water separation process aboard the FPSO involves, among other things, heating 1,074 t/h of wet crude from 50.5°C to 60.7°C. The energy needed to do this is obtained by cooling the 802 t/h of dry crude leaving the separation unit from 87.5°C to 70.1°C.

On board the 416,000-tonne Dalia FPSO are heat exchangers to transfer heat from the dry crude leaving the oil/water separator to the incoming wet crude. Without heat transfer enhancement technology these exchangers would have been four times the size. Photo:Total / Gonzalez Thierry

The designers of this heat exchange system faced a challenge. Available pressures limited them to a pressure drop of 2 bar on the shell side (wet crude) and just 1.5 bar on the tube side (dry crude). Given the high viscosity of the oil – in the range 8–70 cP – this meant that they could not sustain turbulent flow. And as every process engineer knows, laminar flow means poor heat transfer.

Conventional shell-and-tube (S&T) exchangers with segmental baffles and no tube inserts (TEMA type BES) would have required nine units (three sets in parallel, each set consisting of three shells in series), each containing 1,778 tubes with a length of 5.25 m. Together, the nine shells would have had a wet weight of 401 t and required a plot area of 105 m².

Not only was the conventional solution heavy and bulky, but the low fluid velocity on the shell side was also likely to mean poor flow distribution, which in turn would likely contribute to fouling. The end result was likely to have been either even poorer long-term performance than predicted, or frequent downtime for cleaning. Surely a better solution existed?

hiTRAN^® for tubeside enhancement

Good practice in the design of S&T heat exchangers dictates that the heat transfer resistance on the shell side should be the same as that on the tube side. With viscous fluids this is difficult to achieve. Not surprisingly, the conventional design for the Dalia exchangers turned out to be heavily limited by tubeside performance: the heat transfer coefficient (HTC) on the shell side was 455 W/m²K, while on the tube side it was just 95 W/m²K. Improving tubeside heat transfer was clearly vital to better overall performance.

Even in turbulent flow, transferring heat to or from a fluid flowing through a tube has limitations. The problem lies at the tube wall, where friction and viscous forces within the fluid combine to produce a thin layer of fluid that is scarcely moving. The mechanism which carries heat through this “boundary layer” is conduction, which has a low rate for fluids with low thermal conductivity – such as oil.

The heat transfer enhancement technology known as hiTRAN^® Matrix Inserts from Cal Gavin Ltd. solves this problem by increasing the velocity close to the wall, eliminating the troublesome boundary layer. As the illustration shows, streams of blue and red dye cling to the wall of an empty tube (A), but quickly mix into the bulk fluid as soon as they encounter the hiTRAN^® Matrix Inserts (B). Not only does this greatly improve tubeside heat transfer, it also reduces any tendency to fouling.

hiTRAN^® Matrix Inserts quickly move the thin layers of dye from the tube wall into the bulk fluid.

HELIXCHANGER^® for shellside enhancement

Problems with conventional S&T exchangers do not stop at the tube side; on the shell side, the deficiencies of conventional segmental baffles are widely known. A lot of pressure energy is wasted in moving the shellside fluid around the baffles, and a significant amount of fluid is forced through the clearances between the baffles, the tubes and the shell, reducing heat transfer both directly and by distorting the shellside temperature profile. The perpendicular baffles also produce dead spots and recirculation zones where fouling and corrosion can occur.

A better solution is the HELIXCHANGER^® design currently marketed by Lummus Technology Heat Transfer. The HELIXCHANGER^® uses quadrant-shaped baffles to guide the shellside fluid in a helical path through the tube bundle. This creates a flow pattern that is close to plug flow, with uniform velocities and a minimum of bypassing and dead volumes. The result is high shellside thermal effectiveness for a given pressure drop. Industrial experience also shows that the HELIXCHANGER^® can operate for two or three times as long as a conventional exchanger before shellside cleaning becomes necessary.

Instead of turning the shellside fluid through a series of 180° bends, as conventional segmental baffles do, the HELIXCHANGER^® design (shown here during fabrication) guides the fluid smoothly across the tubes, for maximum heat transfer with minimum pressure drop.

A solution for Dalia

Exchangers of the HELITRAN^® design, which combines HELIXCHANGER^® helical baffles with hiTRAN^® Matrix Inserts, proved ideal for the Dalia FPSO duty. The combination of enhancement technologies equalises the discrepancy between the shellside and tubeside HTCs and boosts the overall HTC by a factor of four. Neither of the technologies could have achieved this on its own.

The combination of HELIXCHANGER^® helical baffles with hiTRAN^® Matrix Inserts boosts both the shellside and tubeside HTCs, especially the latter, and removes the disparity between the two. The net result is a fourfold increase in overall HTC.

The revised design provides the same duty as the original (7.9 MW) but requires only two shells. The number of tubes per shell is the same, with a tube length of 6.10 m compared to 5.25 m previously. The plot space required for the HELITRAN^® exchangers is 26 m², or one-quarter of that required originally, and the wet weight of 131 t is one-third of that for the original design (see Table).

Although the improved exchanger design has half the average tubeside velocity of the original, the tubeside residence time is only a quarter of that in the original design, thanks to the improved velocity profile, so the tendency to tubeside fouling is lower. There is a similar effect on the shell side, where the velocity of the helical flow is more than three times the average crossflow velocity in the equivalent segmental-baffle design, and there are no dead spaces. Finally, the use of two shells in parallel means less opportunity for liquid maldistribution – and consequent fouling –compared to the original three shells in parallel.

Nine conventional exchangers (left) are replaced

by two HELITRAN^® exchangers (right)

Even though applying the two enhancement technologies results in a higher cost per shell, the capital cost of the improved design is only one-third that of the original. On top of that, the reduced number of exchanger shells also brings substantial savings in piping and instrumentation.

Compared to many of the advanced technologies required to operate a deep-water field such as Dalia, heat exchangers may seem well-understood or even not worthy of detailed engineering attention. As this case study shows, however, two well-established exchanger enhancement technologies can slash capital costs, space requirements and weight, while also improving operability.

Detailed comparison of the benefits of HELIXCHANGER^® helical baffles with hiTRAN^® Matrix Inserts

Dalia: a deep-water pioneer

With up to a billion barrels of oil, Dalia is one of the world’s largest deep-water fields. Water depths beneath the Gulf of Guinea are in the range 1,200–1,500 m.

Angola’s national oil company Sonangol is the title holder. The operator, Total, owns 40% of Dalia, while StatoilHydro, Esso and BP also have sizeable stakes. The project cost around $4 billion to develop.

Brought on stream in December 2006, Dalia currently produces 250,000 barrels per day and has an estimated life of around 20 years. By 2013 the field will have around 71 wells averaging 3,500 m in length, including 1,000 m of horizontal drilling, in the form of 37 production wells, 31 water injection wells and three gas injection wells.