Jan-2006

Increasing crude unit preheat

Cost-effective exchanger network solutions need to rely on more than just pinch technology if they are to be successful

Scott W Golden and Steve White, Process Consulting Services

Article Summary

Fuel gas prices are increasing and supplies are tight, making energy-efficiency improvements a necessity when revamping crude units. The last time efficiency projects were of major interest was in the early 1980s. During the many low margin cycles since then most of those projects did not meet the corporate investment hurdles of one-to two-year simple payouts. Even so, capital was scarce. Consequently, few were implemented. Refiners are again looking to reduce energy consumption, and pinch technology is once more being touted as the answer. However, the real question remains: does pinch technology address all the considerations needed to reduce crude unit energy consumption, or is it just a mathematical model in the revamp engineer’s toolkit?

Increasing the crude preheat temperature to reduce the fired heater duty and save energy depends on more than just exchanger network configuration. Distillation system design, crude hydraulics, exchanger fouling, fired heater operation and other factors all influence the crude preheat temperature. Exchanger network revamps that achieve a high preheat temperature at start-of-run (SOR) when the exchangers are clean but then limit the crude rate when the exchangers foul, or require specific crude blends to meet the targeted crude charge rate, are not the answer. Pinch technology is no different than other process or equipment models, in that practical know-how must be used to temper one’s answer.

Pinch technology
Linhoff and Vredeveld introduced the term pinch technology to represent a methodology that uses the first and second laws of thermodynamics to identify the minimum energy usage and exchanger network capital costs while recognising the pinch point concept. Pinch technology relies on the stream rates, stream temperatures and the available energy in each heat source stream to predict the minimum fired heater duty, minimum exchanger network area and minimum number of exchanger services.

Pinch technology is good at identifying the minimum temperature approach, yet it ignores crude hydraulics, pumparound (PA) and product pump capacity, exchanger fouling, existing equipment mechanical limits and other peculiarities of an existing crude unit. This makes practical application when it comes to a revamp difficult at best. Non-optimum process flow schemes, poor distillation equipment performance, a high rate of exchanger fouling and unreliable fired heater performance all contribute to a low crude preheat temperature. Heater outlet temperature and pressure, plus the distillation equipment determine the  available heat source duties and temperatures. Low oil velocity or poor exchanger design can lead to rapid fouling, causing the preheat temperature to drop following start-up. When the preheat temperature drops, the heater outlet temperature may also need to be reduced to avoid over-firing. This reduces the distillate yields, which decreases the amount of heat available for the crude preheat.

Pinch technology’s weakness is that the stream flow rates and temperature used in the analysis are dependent on actual process and equipment performance. However, pinch technology specialists who are well versed in the model’s mathematics rarely set foot in a refinery to observe what is really happening. Valid assumptions concerning stream flow rates and temperatures are essential when using pinch models. All critical factors influencing the preheat temperature need to be addressed, not simply the pinch point.

Engineering tools and know-how
Engineering tools speed calculations and allow many alternatives to be evaluated, but do they replace know-how? For example, one recent revamp objective was to increase the heavy vacuum gas oil (HVGO) product yield to unload the coker unit, allowing a higher crude rate. Increasing the HVGO yield requires a higher heater outlet temperature without causing rapid coke formation. The vacuum unit had two fired heaters: the original heater (Figure 1) was started up in 1959 and the second (Figure 2) was added in 2000. Engineers using fundamental principles and slide rules designed the original, while the heater installed in 2000 was designed with the latest models including computational fluid dynamics (CFD) software. The heater built in 1959 operated for four years without requiring decoking at a radiant section average heat flux of 10 500 btu/hr-ft2-°F, but the new heater had coking problems at 9000 btu/hr-ft2-°F.

Both heaters are four-pass single-cell box-type heaters operating without any coil steam. The 1959 heater was designed with two passes on each wall using serpentine tubes, with pass outlets exiting the top of the box. The new heater had the passes stacked on the walls — one pass on each wall was upflow and the other downflow — with the outlet tubes exiting the middle. Both heaters were operated with pass flow rate balancing to maintain the same outlet temperature from each pass. The original heater operated with nearly equal flow to each pass, while the new heater required a high flow rate in the two passes located in the lower section of the heater and a low flow rate in the upper two passes. The original heater had equal heat absorbed per pass (equal heat flux), whereas the new heater had a high heat flux in the bottom passes and a low heat flux in the passes in the top of the radiant section. Furthermore, the original heater operated at mass flux rates of 350 lb/sec-ft2 in each pass, while the new heater operated at 250 lb/sec-ft2 in the high flow rate passes in the lower part of the radiant section. The new heater has had two-year run lengths, while the original heater has operated for four years between decoking since 1959.

The first heater design is fundamentally sound and the second is not. The revamp engineer has the latest software (which was used to design the new heater) at his disposal, yet the original design using slide rule calculations and fundamental principles, such as balancing the “pass” heat flux, minimising the radiant section heat flux variability and optimising the oil mass flux (to reduce the rate of coke formation), was better. Fundamental principles and lessons learned from other operating heaters went into designing the original heater, but the new heater was designed by modelling specialists relying on complex computer models and failing to consider fundamental principles and lessons learned.

Crude unit heat exchanger network design is more complicated than heater design. As previously mentioned, it involves crude hydraulics, distillation, heat exchanger and heater system performance. Computer models are great tools, but there is no substitute for know-how. Successful application of pinch technology, like other models, depends on users’ experience of the unit being revamped. Mathematical model results alone will not produce a reliable crude unit heat exchanger network.