Enterprise Product Partners LP expanded its NGL fractionation plant in Norco, La., to feed 75,000 b/d to its de-ethanizer and 90,000 b/d of raw feed (less ethane) to its depropanizer. After expanding the plant in 1997 for a throughput of 70,000 b/d of relatively lean NGL feed with a propane and heavier capacity of about 42,000 b/d, Enterprise anticipated higher flow rates and a much heavier NGL feed.
The expanded plant started up in November 2003. Because of increased heat integration, the utility efficiency of the plant decreased about 15% in terms of btu/gal of feed.
The new expansion uses more-extensive heat integration to lower utility costs and, where possible, the new design used existing equipment and infrastructure to minimize investment cost.
The main changes for the expansion were the conversion of an existing depropanizer to a high-pressure debutanizer and the addition of a new depropanizer column and a second heat-pumped deisobutanizer.
The high-pressure debutanizer rejects heat into a new side reboiler on the existing de-ethanizer and reboils the existing deisobutanizer.
This article focuses on the decision process used in the design’s evolution.
Pre-expansion facility
The Norco plant fractionates a full range NGL feed into ethane, propane, butanes, and natural gasoline. Both feed and product are transported via major pipeline systems. Gas plants located in the region provide the raw feed.
Before the 2003 expansion, the plant was designed to process 70,000 b/d of raw NGL feed with a propane and heavier content of 42,000 b/d. The feed, however, was not as lean as expected and, in 2002, the effective plant capacity was 55,000 b/d.
The original plant design included hot oil as a heat source for the fractionation column reboilers and cooling water for column overhead vapor condensing. Some heat integration was implemented in the 1997 expansion to the plant by use of the debutanizer overhead as a heat source to reboil the deisobutanizer (Fig. 1).
There were also some feed-product exchangers that allowed some of the hot product from the debutanizer, depropanizer, and deisobutanizer to exchange heat with the raw feed.
Expansion objectives
To handle the additional raw NGL streams, Enterprise needed a design to expand the Norco fractionation plant to a raw de-ethanizer feed of 75,000 b/d. In addition, the new equipment should have a capacity of 90,000 b/d. Finally, Enterprise wanted increased energy efficiency to meet standards set at its other facilities.
Plant capabilities, constraints
First, Enterprise rated its existing equipment. The de-ethanizer column showed a capacity of more than 80,000 b/d with existing trays and with the addition of side reboilers.
The depropanizer and debutanizer were undersized and would have to be paralleled. It was clear that a second deisobutanizer column would be required because the existing column could process only about 50% of the available feed.
Three new columns would therefore be needed if the existing process flow were retained; however, the existing depropanizer’s capacity, when placed in debutanizer service, was sufficient for all of the debutanizer feed that the expansion required. The column’s pressure rating was high enough that it could handle the high-pressure debutanizer service and heat from condensing the column overhead could be used.
The existing depropanizer, when placed in debutanizer service, had an NGL feed capacity of more than 90,000 b/d. A new depropanizer column was needed for this change, but it also could be sized for 90,000 b/d.
The existing debutanizer column was too small for any of the new capacity requirements and was subsequently left idle.
One of the original design concepts for heat integration included a new high-pressure deisobutanizer so that the additional heat could reboil the existing deisobutanizer. Heat from the existing debutanizer overhead would then be available to reboil the de-ethanizer through proper placement of a side reboiler.
An order-of-magnitude estimate was prepared based on the equipment list for the changes previously mentioned and in Fig. 2. The estimate indicated a reasonable return for the expansion.
Although the screening study appeared to optimize the column configuration, a comparison to other Enterprise facilities indicated that additional energy efficiencies would result from further design development.
Design evolution,
Enterprise conducted optimization studies to determine the heat-integration strategy that would best match the available supplies of heat with the heating requirements.
These optimization studies helped determine what portion of available heat from the debutanizer overhead could be absorbed by the de-ethanizer. If excess heat could overload the column’s vapor-handling capacity, other heat requirements would have to be identified to utilize the remaining heat.
Because the original proposal heat-integrated the new deisobutanizer with the existing one, the new column could possibly be reboiled with heat from the debutanizer. This possibility had to be tested to see if it resulted in excess heat.
Another alternative was to heat-integrate the debutanizer overhead with the reboiler requirements of both deisobutanizers. This could consume all the heat from the debutanizer overhead, but it was unclear whether additional heat from the hot oil system would be needed. If hot oil were needed, it could provide a justification for using a heat pump to supply the heat for the new deisobutanizer and split the heat available from the debutanizer between the existing deisobutanizer and the de-ethanizer.
Using debutanizer heat for the de-ethanizer via a side reboiler could provide extra capacity in the de-ethanizer column by reducing the liquid and vapor traffic in the column’s bottom. This could eliminate the need to retray the column for the new feed rate.
Economic considerations
We wanted to maximize the use of available heat to save energy and to reduce the capital investment required for the project. Using all the heat available from the debutanizer overhead could reduce the size of the new hot oil furnace and cooling tower required in the expansion.
Another significant reduction in the hot oil and cooling water systems was also possible if the new deisobutanizer was equipped with a heat pump. The tradeoff would be the additional cost of the compressor and additional energy consumption associated with its driver. This tradeoff would necessitate an economic evaluation against a logical base case.
Debutanizer studies
Enterprise evaluated four different evolving designs to use heat from the debutanizer overhead: integrate with de-ethanizer side reboilers, integrate the new and existing deisobutanizers, baseload the new and existing deisobutanizers, baseload the existing deisobutanizer and heat integrate the de-ethanizer side reboilers.
De-ethanizer side reboilers
Heat available from the debutanizer overhead could be heat-integrated with the de-ethanizer using side reboilers.
This concept required raising the debutanizer’s operating pressure to increase the overhead temperature. This would ensure a temperature approach between the cold-side outlet and hot-side inlet of the side reboiler of at least 5° F.
When the debutanizer’s operating pressure was raised to within 15% of the set pressure on the relief valve, the maximum heat to the de-ethanizer column was achieved. Further attempts to increase the heat duty to the side reboilers decreased the approach temperature to less than the 5° F. required.
To help maximize the heat recovered, the duty was distributed between an upper and lower side reboiler.
The trap-out trays for these reboilers were located at Stages 3 and 7 from the bottom of the column. After several attempts to optimize the duty and location of the reboilers, we found that only about 65% of the available heat from the debutanizer column could be used in the de-ethanizer; the rest would have to be condensed against water or air.
New to existing deisobutanizer
An examination of the overhead vapor from the proposed high-pressure deisobutanizer column, when heat integrated with the existing deisobutanizer reboiler, indicated that 15% of the overhead duty would need to be condensed against air or water. This resulted from the need to increase the tower pressure to maintain a reasonable approach between the condensing inlet on the high-pressure deisobutanizer and the reboiler outlet on the low-pressure deisobutanizer.
It became apparent that considerable heat was rejected to the atmosphere as a result of the heat imbalance between the debutanizer and high-pressure deisobutanizer. It was becoming apparent that one or both of the deisobutanizers would have to be a heat sink for the debutanizer overhead.
In addition, the tray loadings were much higher when the debutanizer column operated at a higher pressure. The tray supplier determined that the maximum vapor and liquid loadings for the existing trays were more than 20% above their rated capacity. New trays would be required for the column to run at the higher pressure.
Baseload deisobutanizers
One alternative for full heat integration of the debutanizer overhead was to baseload the reboilers of the new and existing deisobutanizers. Fig. 3 shows this concept with the heavy lines.
This design provided two major advantages over the original proposal. The first was that it allowed full use of the heat from the debutanizer condensers.
The second advantage was that it allowed the debutanizer and the new deisobutanizer to operate at lower pressures. This could potentially reduce the new equipment cost by reducing the vapor-liquid traffic in the column, which eliminated the need to retray the column.
In addition, it lowered the reboiler duty requirements of the debutanizer and deisobutanizer by more than 25 MMbtu/hr. This reduction represented more than 23% of the additional heat duty required by the original project proposal. A disadvantage was that it did not fully reboil the new deisobutanizer column. Excess heat from the debutanizer would supply less than half of the heat requirement. That left about 20 MMbtu/hr needed from the hot oil supply. Due to the heat reduction provided by lowering column pressures and fully integrating the debutanizer overhead duty, however, the total reduction was approaching 43% of the new heat duty to be provided.
A second disadvantage of integrating all of the heat from the debutanizer overhead with the deisobutanizer columns rather than the de-ethanizer was that side-reboil on the de-ethanizer reduced the liquid loading at the column’s bottom. The higher liquid loads increased the likelihood that the column would have to be retrayed at base design rates of 75,000 b/d.
Although Enterprise recognized that the column would need to be retrayed at 90,000 b/d of throughput, this design could possibly use existing trays if side reboil were used.
To reduce the disadvantages of baseloading the deisobutanizer columns with debutanizer heat, Enterprise developed a proposal to use the heat to side reboil the de-ethanizer and the existing deisobutanizer.
This could potentially defer the column retray and would allow all of the debutanizer overhead duty to be heat integrated into the process. It also allowed the possibility of using a heat pump on the new deisobutanizer column, if justified, rather than a conventional hot oil reboiler.
Final design
If the new deisobutanizer overhead was not heat integrated with the existing deisobutanizer, a lower operating pressure was possible. The hot oil, however, would have to supply a reboiler duty of about 40 MMbtu/hr.
If the column was reboiled with heat from the discharge of a compressor that would handle the overhead vapor, all the heat could be supplied from the heat of compression and condensation of the overhead. This would potentially reduce the hot oil requirements by a corresponding amount.
With a heat pump on the new deisobutanizer, baseloading the existing deisobutanizer, and heat integrating the de-ethanizer side reboilers reduced the hot oil requirements by 70% compared to the original proposal. The cooling-water requirement was also reduced and overall energy savings of 25% were possible. Fig. 4 shows the final process flow scheme.
Heat pump economics
Because the heat pump would add to the project’s capital cost, Enterprise conducted an economic evaluation to show that a reasonable return was possible. The heat pump would increase the equipment cost $327,000 and the total installed cost about $569,000 compared to a conventional deisobutanizer operating at low pressures and condensing against cooling water.
Table 1 shows all the equipment costs.
The main elements were an additional heat pump with larger reboiler and subcooler for the heat pump process. The conventional column had a smaller reboiler but additional condensing surface area. The total costs of exchanger surface area were roughly equal. There were higher costs for cooling tower and hot oil heater capacities for the conventional deisobutanizer. The heat pump lowered the hot oil heater capacity to 32 MMbtu/hr from 71 MMbtu/hr of heat absorbed. The cooling tower was about 30% smaller.
The energy savings resulted from obtaining low-cost electricity at $0.065/kw-hr vs. buying natural gas at $3.50/MMbtu for firing the additional furnace capacity. The effective energy savings was $996,000/year. The heat pump compressor paid for itself in less than 7 months with these utility savings and investment costs.
De-ethanizer capacity
An initial investigation of the de-ethanizer revealed that, at the lower feed rate of 75,000 b/d, we could introduce combinations of feed preheat and side reboiler duty to allow the column to operate without a retray. This helped define the de-ethanizer as an important point of heat integration in the overall process.
The de-ethanizer’s rectifying section would experience more vapor and liquid traffic when the preheat and purity of the ethane in the overhead increased. Higher ethane purity is attractive due to the economics of upgrading the propane value from that of ethane. From an economical standpoint, minimum propane in ethane is desirable up to the vapor-handling capacity of the column’s rectifying section.
Increasing the preheat could also reduce liquid traffic in the column’s bottom. Earlier assessments of tray capacity indicated this effect.
It was also more economical to operate with maximum preheat to reduce the amount of hot oil required for the bottoms reboiler. The feed preheat could therefore potentially avoid the need to retray the bottom section if the top section’s capacity would allow it.
Side reboil also helped reduce liquid traffic in the column’s bottom but increased the load to the rectifying section. When the preheat decreases due to a colder incoming feed, side reboil duty can increase to compensate for the reduced loading in the stripping section.
De-ethanizer tray capacity
The de-ethanizer had high-capacity trays installed during the previous expansion to 60,000 b/d. These trays were rerated based on loadings obtained from a case run at 90,000 b/d and a relaxation of the propane specification in the ethane product. The existing trays could handle a feed rate of 82,000 b/d.
We addressed the tray capacity as related to the propane-in-ethane specification for the 75,000-b/d case and recognized that some spare capacity might be available.
Based on the vendor ratings for the existing trays, we simulated various combinations of feed preheat, side reboiler duties, and propane content in the ethane at 90,000 b/d of raw feed to the de-ethanizer. Attempts to reduce the vapor load in the column’s top were successful at the higher propane content selected for the original proposal.
The case runs resulted in a maximum preheat temperature of 90° F. Liquid loading in the column’s bottom was beyond the capacity of the existing trays. Attempts to raise side reboiler duty to mitigate this limitation were only partially successful as maximum liquid loading was shifted from the bottom of the column to just above the upper side reboiler draw.
Tray replacement would be required at higher feed rates and, therefore, we returned the propane specification to its original level in the overhead to determine the tray loadings for the new design.
With the existing trays it was possible to balance feed preheat against side reboiler duty and obtain a propane specification in the ethane that met the original requirements at the design rate of 75,000 b/d of raw feed. Enough liquid handling capacity was available in the de-ethanizer rectifying section trays to even allow a lower propane specification than the original requirements without exceeding either vapor or liquid-loading limitations.
De-ethanizer retray justification
A de-ethanizer retray was not needed to achieve the design rate of 75,000 b/d as long as side reboilers were installed to help unload the liquid handling limitations in the column’s bottom. This, however, required modification of the existing trays to include trap-out trays for the two side reboilers that were considered as possible candidates for heat integration.
Because the column would have to be vapor free, the column retray for a rate of 90,000 b/d was a good idea. This would avoid any internal modifications to the de-ethanizer for future expansion programs and we considered it a prudent investment.
Depropanizer conversion
The existing depropanizer column was a conventional fractionation column with overhead vapor condensed against air and bottoms reboiled with hot oil. The column was equipped with 40 high-capacity trays.
As originally proposed, the column would be converted to a debutanizer. Available heat from condensing the overhead vapor would side reboil the de-ethanizer. This required the converted column to operate at its upper pressure limits. In this operating region the trays would limit the column’s capacity to less than the design basis requirement due to higher internal reflux. The de-ethanizer could not efficiently use all of the heat available. If the existing deisobutanizer were used for most of the heat integration, the debutanizer pressures and tray loadings would be lower. This would allow us to use the existing trays for the new service.
As a debutanizer, the column’s bottom would operate at a higher temperature than as a depropanizer. For the new conditions the reboiler would limit significantly the amount of heat that could transfer with the existing surface area. To remove this limitation, we installed an additional reboiler in series, which increased the total installed surface area.
We ultimately completely heat integrated the debutanizer overhead with the existing deisobutanizer and de-ethanizer side reboiler in the new configuration. This alternative uses the existing heat-integration reboiler on the existing deisobutanizer and provides heat to a new side reboiler for the de-ethanizer.
Existing air-fin coolers on the existing depropanizer were not required for the new service, but were left available for start-up.
Process flexibility
The design pressure of the existing column was significantly higher than required for its operation as a debutanizer. This gave the column more flexibility to accommodate heat integration for a wider range of raw feed compositions.
Increasing the debutanizer operating pressure to maintain an adequate temperature differential for good heat transfer accommodated the higher de-ethanizer bottoms temperature due to compositional changes. The existing air-fin exchangers, available for condensing overhead vapor, would facilitate column start-ups. With the current design, however, all of the heat associated with the overhead vapor could integrate into available requirements.
Grassroots depropanizer,
Originally, the new depropanizer had a design with 45 trays and a 13-ft ID. More trays would increase the costs. Less energy consumption associated with reduced internal reflux requirements for the larger number of trays would offset these higher investment costs.
Adding 15 trays reduced internal reflux and decreased the column diameter by 1.0 ft. The trays are spaced at 24 in., therefore, the overall column tangent-to-tangent length increased 30 ft. The total tower cost increased $193,000. Other equipment was not affected much and the total differential cost for an increase in tray count was $197,000.
The main reason for the economic evaluation was to reduce hot-oil consumption. The 15 trays reduced the heat consumption slightly less than 6.0 MMbtu/hr of absorbed heat. The fuel savings were $239,000/year based on fuel cost of $3.5/MMbtu. This provided shorter than a 10-month payout and made our decision to increase the column tray count to 60 an easy one. ✦
The authors
Robert Moss ([email protected]) is the director of technical services for Enterprise Products Partners LP, Houston, with responsibilities for process engineering, pipeline optimization, and mechanical reliability. He started with Enterprise at Mont Belvieu in 1993 and has more than 20 years of industry experience. Moss holds a BS in chemical engineering from Texas A&M University.
Gene Flipse ([email protected]) is process engineering manager for San Juan and Western operations at Enterprise Product Partners LP, Houston. His previous positions inlucded process engineering roles with Union Carbide Corp. and Celanese Corp. He holds a BS in chemical engineering from Virginia Tech, and an MS in chemical engineering and a doctor of engineering from Texas A&M University..
W.T. Skip Chandler is a senior project manager for URS Corp., New Orleans. He previously worked for Marathon Oil Co. Chandler joined URS in 1988 and has worked as a senior mechanical engineer, construction manager, mechanical/piping department head, and engineering manager. His recent experience has focused on projects in the midstream NGL market and he also has project experience with refineries, pipelines, and onshore and offshore oil and gas production facilities. Chandler holds a BS (1978) in mechanical engineering from Tulane University, New Orleans..
Ed Orloski is a senior process consultant with URS Corp., New Orleans. He has worked extensively with the process design and optimization of gas and hydrocarbon liquids processing in both upstream and downstream facilities. Since joining URS in 1998, he has actively pursued cost-benefit analysis of processes associated with treatment, transportation, and processing of gas and liquids production for onshore and offshore applications. Ed holds an MS in engineering from Oklahoma State University.