NOVEL C3-RECOVERY SCHEME INSTALLED IN TEXAS GAS PLANT

Fred S. Crum, James K. Dyer
Mitchell Energy Corp.
The Woodlands, Tex.

Mitchell Energy & Development Corp., The Woodlands, Tex., has installed in a Southeast Texas plant a novel gas-processing design specifically to permit propane production.

Rejecting two more conventional process designs, Mitchell Energy settled on refluxing design, built a 5,000-b/d unit near Bryan, Tex., and has received a U.S. patent for the design.

NO PROPANE PRODUCTION

In 1988, Mitchell Energy began to study the economics of producing propane in the Bryan, Tex., area where it operated two cryogenic gas-processing plants, each producing a demethanized raw mix product.

The Bryan-Hicks plant was built in Brazos County near Bryan and installed in December 1982. It is jointly owned by Mitchell's Liquid Energy Corp. unit (45%) and UPRC, Ft. Worth (55%). (The new fractionator is owned 100% by Mitchell.)

The Madison plant (three cryogenic units) is about 20 miles northeast in Madison County and was installed in January 1989 (Fig. 1).

Nominal gas-inlet capacity at the Bryan plant is 20 MMcfd; current throughput is 12.6 MMcfd. The combined Madison plants can process up to 60 MMcfd (20 MMcfd each); current combined throughput is 37 MMcfd.

Combined production from the two plants has been about 180,000 gpd: Bryan 68,000 gpd, Madison 112,233 gpd.

Before the project, product from the Madison plant moved through pipeline to the Bryan plant for treating (CO2 removal) and was then delivered to a Koch product pipeline for transport to Mont Belvieu, Tex., for final fractionation.

At that time, there were no propane-producing facilities near Bryan. Propane sold in the area arrived in trucks from Mont Belvieu or other sources more than 100 miles away. Additionally, there is no market for ethane around Bryan.

An earlier study had shown propane production to be uneconomical on smaller scales of operation and utilizing conventional fractionation processes. When the Madison County plant came on stream and thereby increased the volume of feed available for fractionation, the study was revived.

Table 1 shows the ideal material balance around the proposed fractionator. The feed consists principally of ethane, propane, butanes, and pentane plus with small amounts of methane and carbon dioxide.

Mitchell Energy wanted to remove the propane for local sale and to return the remainder of the stream to the pipeline for transport to Mont Belvieu. Solving this relatively simple fractionation problem involved installing two fractionating towers arranged as shown in Fig. 2 (which for simplicity omits the condensers and re-boilers).

The net overhead from the first tower is ethane, the second tower overhead the propane product, and the bottoms the butane and heavier fraction.

The de-ethanizer overhead would then be recombined with the depropanizer bottoms for delivery to the product pipeline. The propane would be stored for sale off the truck rack.

DISADVANTAGE

This effective and often-used process has only the disadvantage that the condensing temperature on the de-ethanizer is in the range of 30-400 F.

The condensing temperature is determined by the operating pressure of the column, which is limited by the critical pressure of the bottoms product of about 560 psia.

This limitation prevents the tower pressure being raised high enough to condense the reflux with either air or water. Therefore, refrigeration must be used to condense the de-ethanizer reflux.

Also, a higher operating pressure will require more reflux for a given separation since the relative volatility of the key components decreases with increasing pressure.

In an alternative fractionation scheme (Fig. 3), the net overhead from the first tower is ethane and propane. Operating the first tower at 445 psig permits the reflux to be condensed at 1200 F. with air. The second tower requires refrigeration to condense reflux however.

In yet a third scheme (Fig. 4), the ethane is removed from the feed in the first tower. The C3 and C4 + are fed to the second tower where the C3 goes overhead and the C4 + is the bottoms The depropanizer bottoms are then recirculated to the top of the de-ethanizer.

The net effect of this scheme is that refrigeration is no longer required to provide reflux on the de-ethanizer. De-ethanizer reflux condensing can be accomplished with air.

The net overhead product from the de-ethanizer is the C2 and C4 + which was to be sent to the product pipeline. The C4 + reflux on the de-ethanizer prevents the C3 from leaving the top of the de-ethanizer.

This scheme looked promising because an entire system (refrigeration) would be eliminated from the process.

Figs. 5, 6, and 7 illustrate the compositional changes taking place- within the first tower of each process.

Fig. 5 shows the tray liquid compositions for the first process considered. In this process, the C, went overhead and the C3 and C4 + went to the bottoms.

Fig. 6 shows the tray liquid compositions for the second process considered. In this process, both the C2 and C3 went overhead and only the C4 + went to the bottoms.

Fig. 7 shows the tray liquid compositions for the third process considered. In this process, which has come to be called the new process, C4+ is present throughout the tower.

The C4+ reflux absorbs most of the C3 vapor so that very little C3 goes out the overhead.

PROCESSES COMPARED

A detailed simulation was prepared for each process. Table 2 shows a comparison of equipment sizes required for each process.

The new process requires a larger de-ethanizer than either of the other two processes because of the recirculation Of C4+ through the de-ethanizer which does not occur in either of the two other processes. The depropanizer size is not significantly affected by the C4+ recirculation.

Table 3 shows the reboiler and condenser duties for the three processes. The total reboiler duty for the new process is significantly greater than for either of the other two processes.

Table 4 shows the power requirements of the three processes. The new process requires more pump and air cooler fan horsepower but much less gas-engine power, assuming that refrigeration compression would be gas-engine driven.

Table 5 tabulates the fuel requirements for the three processes. Total fuel required was less for the new process primarily because no engine fuel was needed.

In terms of total power and fuel costs, the new process uses more electricity but less fuel than the other two processes (Table 6). Since electricity is more expensive than fuel on a unit-of-energy basis, the processes become more even in terms of total operating cost.

The new process, however, is slightly more economical in terms of operating cost. Conventional Process No. 2 is too expensive because of its high energy requirements.

A comparison of capital costs shows Conventional Process No. 1 to be about $400,000 more expensive than the new process (Table 7). The capital cost of Conventional Process No. I was arrived at by adjusting the new process cost for differences in equipment items and sizes.

INSTALLED, PATENTED

In the new process (Fig. 8), feed enters the de-ethanizer which is operated at 370 psig. Depropanizer bottoms are mixed with the de-ethanizer gross overhead and cooled to 1200 F. At these conditions, the stream is still two phase.

The net overhead vapor is compressed and net overhead liquid pumped to 525 psig at which it totally condenses in the feed exchanger.

The de-ethanizer overhead could have been totally condensed if the C2 content of the feed had been lower, the tower operating pressure higher, or more C4+ recirculated. The total process heater duty was reduced by 15% by addition of the side reboiler on the de-ethanizer.

Table 8 shows a detailed material balance which was identical for all three processes compared. Propane product recovered is 94% of that contained in the feed. Product treating for C02 removal is performed downstream of the propane fractionator.

Mitchell Energy concluded that the new process was the best choice for the Bryan application.

The new process is particularly well-suited to the market for propane. In cold weather when propane demand is highest, the condenser temperature of the de-ethanizer can be lowered which has the following benefits: