BASIC TECHNOLOGY CAN IMPROVE REFINERY PROFITS

Norman P. Lieberman
Process Improvement Engineering
Metairie, La.

A set of 12 principles based on refinery troubleshooting experience emphasizes the application of basic engineering methods to improve the design, retrofit, and optimization of refinery process equipment.

PROCESS PROBLEMS

At a recent social gathering, several retired colleagues, many of whom had been process engineers during World War 11, were telling stories about their refinery careers. One colleague related an incident that occurred in El Dorado, Kan., in 1948, that was almost identical to a problem recently encountered by the author'

As the string of stories from an era when pneumatic control was new grew long, it became apparent that nothing had changed. The refinery process problems that undermine profitability today are the same ones these industry veterans encountered in their youth.

PROFIT BOTTLENECKS

Every chain has a weak link. Every process unit has some component that bottlenecks throughput. This also is true for refinery profits. There is always a component that is bottlenecking profits.

Building and operating a refinery requires several critical components:

Land
Labor
Capital
Raw materials (crude oil)
Electric power
Water
Combustion air
Administrators
Business expertise
Legal guidance
Technology.

While air is readily available and capital can always be had for a price, our industry is suffering from a severe shortage of technology.

Every refinery manager who has ever watched a flare burn up profits has known that the key to increased profitability is to use technology to improve process operations.

But what sort of technology is required?

Do we need more of the same, such as doubling the technical department next year? Or do we need new and better technology, such as implementing a more advanced computer control system?

NEW FACE, OLD FACE

Refineries have two faces. Their new face is a mosaic of improved catalysts for cracking, reforming, and isomerization, plus enhanced computer control strategies. Their old face is a vast array of distillation towers, heat exchangers, vapor-liquid separators, settling drums, centrifugal pumps, steam ejectors, storage tanks, reciprocating compressors, steam turbines, and other operating equipment, all of which are subject to poor design and inconsistent operation.

It is this collection of ordinary machinery, that generated the great majority of lost profits in a petroleum refinery.

The bulk of the process equipment in a refinery is pretty much the same today as it was in the 1940s. That is why the retired engineers related stories similar to today's experiences.

There is no significant way to make a sieve tray better, and a floating-head, shell-and-tube exchanger is still the state-of-the-art method of transferring heat between process fluids. The Woodward Governor on a steam turbine is a great improvement over the hand-speed control valve, but this innovation was introduced decades ago.

A MATURE INDUSTRY

What sort of approach, then, is needed in an industry employing equipment that has largely reached a mature level of development? Probably the revival of the type of technology used to develop the original equipment.

This technology requires the application of:

Ideal Gas Law
Bernoulli's equation for hydraulics
Vapor-liquid equilibrium (Rauolt's, Dalton's, and Henri's Laws)
Sensible and latent heat balances
Stokes' Law for settling velocities
Material balances (converting barrels to moles)
Heats of formation (free energy).

This has always been well known. The difficulty is in achieving the application of this basic technology at the plant level.

Why is it so difficult, then, especially for major refineries, to use basic technology to debottleneck refinery. profits? There are three reasons:

Young engineers receive extensive university training on hi-tech subjects, but have forgotten the basics by the time they are hired as refinery engineers.
Older engineers become enmeshed in the bureaucratic procedures of the refinery and neither train the young engineers in basic technical skills nor apply such skills themselves.
Senior engineers enter management and are removed from the basic and vital engineering function.

LOST REALITY

In one example of using common technical sense to solve operating problems, Process Improvement Engineering was working with a young technical service engineer to improve the operation of a ballast-water benzene stripper. The problem was that no benzene was appearing in the offgas vent-ed from the reflux drum (Fig. 1).

When asked how he controlled the reflux drum temperature, the engineer said he added enough stripping steam to maintain a temperature of 180 F. in the reflux drum. This, he said, would ensure that any benzene stripped overhead would remain a vapor in the offgas. The reflux drum, however, was cold to the touch - about 80 F.

The engineer was then asked which temperature he thought was correct, 80 F. or 180 F. (which had been measured 200 ft. downstream of the reflux drum after flowing through an Sin., steam-traced vapor line). Even though his hand was on the reflux drum, he said he did not know and would have to check.

SCIENTIFIC METHOD

All too often, discussions and decisions reached by group consensus are based on faulty data or theoretical considerations. The lengthy meetings often required to reach such decisions are frequently unnecessary.

The problem is that the participants do not field-test their theories before meeting.

In the case cited previously, if benzene vapors were not leading with the offgas and the stripped ballast water was free of benzene, the benzene must have been accumulating somewhere (probably in the reflux drum). But the reflux drum was thought to be too hot to condense the benzene vapors.

Does this mean that the technical manager, who guessed that the drum was not 180 F. should have climbed up to the reflux drum and placed his hand on it? Or should he have taken the more convenient route of asking the unit engineer to have the instrument department calibrate the temperature indicator shown in Fig. 1 ?

Decisions driven by data enhance profits. Decisions derived from opinions' process models, and speculation waste capital and expenses. The reality of process operations is revealed by contact with the process equipment. It does not appear on the computer screen.

The process of making decisions based on field data is simply the application of the scientific method. The truth of a theory is tested by observation, measurement, and experimentation, when possible. When a measurement or an experiment is not possible, calculations are made.

To debottleneck refinery profitability, basic technology must be employed. And the most basic of all technology is the scientific method, which means that decisions must be based on reality, opinions count for nothing, and experimentation is supreme.

APPLYING TECHNOLOGY

Profitability in a refinery comes from the technical person's analysis of the information gathered through the operators and process equipment. Many clients have asked how this mental asset can be enhanced.

Through many years of process troubleshooting, the author has developed 12 principles to improve the application of basic engineering technology to the design, retrofit, and optimization of refinery process equipment:

Begin change at the top. Operations and technical managers must be technically competent. If the pressure drop across a crude tower is 38 psig and the specific gravity of the liquid in the tower is 0.75, the height of liquid in the tower above the flash zone is at least 120 ft. If the entire height of the crude tower is only 100 ft, the reported data cannot represent the real situation. Even the plant manager should be able to determine this from the data.
Make leadership a priority. junior engineers only work as practical field engineers if then, see their supervisors in this role. Does this mean that a technical manager who suspects a pressure drop survey on a crude tower to be flawed should climb to the top of the column to verify the location of the top-pressure sensing point? Making such an appearance in the field demonstrates that even a senior manager is capable of pulling a thermocouple to verify a critical temperature, and displays the sort of action management expects from the technical staff.
Know your refinery. Process equipment is more than an image on a screen in the control room - it is real. Everyone on the management and engineering team should walk through the process units with a flow-sheet, locate the towers and heat exchangers, touch the pumps and control valves, and look into every firebox.
This exercise should bring up questions, such as: Is it normal for the wet gas compressor to be located 1,500 ft from the crude unit overhead receiver? Why does the forced-draft air blower on the crude heater look like it has not operated in 20 years? Shouldn't the air intake on the fluid catalytic cracking unit blower have a bug screen in front of the intake filters?
Plan turnarounds. Half of the opportunities to implement technical improvements to process units can be done only when the unit is down for a turnaround. Turnaround planning should begin the day after start-up. If a vacuum tower is 1 month into the run and the light vacuum gas oil (LVGO) product is observed to be mostly diesel, check the delta P across the crude-tower bottom stripping trays. If the delta P is zero, the trays are damaged.
Ask and answer the following questions: What sort of technical improvements can be made to enhance the mechanical integrity of the trays? What is the calculated pressure drop of the stripping vapor through the sieve holes? How does varying the steam rate affect the distillation of the LVGO?
Continued plant testing to form a basis for the next unit turnaround should be a primary function of the unit technical service engineer.
Support and protect your technical staff. Good technical work should be viewed as critical to the debottlenecking of refining profits. Do not waste this resource by using the technical staff as a conduit to supply day-to-day operating information to management. Management should obtain this information from daily lab reports and direct discussion with shift operators.
Good technical work stems from hands-on data collection. The shift operators must be satisfied that an engineer has to perform certain tests to do his or her job safely and properly. The operators, therefore, must not interfere with engineers catching samples, installing pressure gauges, changing dial thermometers, or performing other such work.
If the shift operators interfere with this activity it becomes a problem for management. Do not expect a young engineer to stand up to a veteran operator without management support.
Encourage written communications. Technical information is best transmitted in the form of pictures elaborated by numbers. Verbal communications, or meetings, are an excellent way to form intergroup bonds between employees. They create a real opportunity to exchange original ideas. But they are a poor way to transmit technical information.
A technical meeting should be preceded by a written technical report. Ideas or questions arising from the report should be submitted to the author for evaluation before the meeting. The attendees should read the report before the meeting. The purpose of the meeting is to determine the action to be taken from the facts presented in the report.
Written technical reports should be filed and historical files should be maintained describing technical work on process units.
Run performance tests. Performance tests are like an annual physical exam for the unit. They force the engineer running the test to learn his unit and summarize, in a single document, all the opportunities for improved profitability. Such tests provide an historical record of the unit operation for future process engineers.
The performance test report is a vehicle for the application of technology in a process plant. The entire technical staff of the refinery can study this "snapshot" of the process. After running a performance test, for example, a utility engineer might see that the cause of his refinery's chronic steam imbalance is the use of a topping turbine that he had been told was never run.
Define the purpose of technical management. The primary job function of the technical manager is to assist young engineers on problems they are unable to solve themselves. This is best done by working with them in the field to obtain data, and by illustrating how calculations are performed.
The primary purpose of refinery engineers is to observe, measure, and calculate the operating parameters of the process units. Because many technical managers spend most of their time interfacing with upper management, experienced technical personnel in the refinery have little opportunity to work directly with the process equipment.
If the senior technical staff communicates less with upper management, the decision-making process will be pushed down the organizational ladder, as it should be.
Insist on technical competence in the engineering staff. While the level of technology needed to design and optimize refinery process equipment is basic, competence in this basic technology in university chemical engineering graduates is far from universal. The problem is bad and getting worse.
The typical new process engineer often must be retrained in engineering basics. This is best done by reading and working through sample calculations. A good way to establish that an acceptable level of technical competence has been reached is through written testing. (For example, have the engineer solve problems such as determining the dew point of a mixture of 80% propane and 20% isobutane at 200 psig.)
Supply the tools to do the job. If the engineering staff is expected to spend 50% of its time gathering data, they ought to have the right tools. A well-equipped tool box contains:
- Pipe and crescent wrenches
- Wheel keys
- Every fitting known to mankind
- Digital pressure gauges with extensions
- Differential, battery operated, digital pressure transducers
- Infrared temperature indicators
- Differential digital surface temperature indicators
- Vacuum mercury manometer
- Portable microsample cooler.
The manual skills needed to safely perform pipefitting tasks incidental to technical work ought to form part of the training of process engineers.
Temper the demands of upper management. A tremendous percentage of technical time is wasted responding to questions from upper management. Middle-level managers should use their engineering judgment to give an approximate answer. It is better that the technical service manager spend 2 hr generating a reasonable estimate of reformer yields than the unit engineer spend 2 weeks correlating a year's worth of data.
Make use of errors. Mistakes are a sign of people working. The author recently popped the safety pressure-relief valve on a client's crude tower. The purpose of this action was to pinch off the inlet to the overhead condenser to gauge the effect on the cooling water outlet temperature.
It was an ill-conceived experiment that accomplished nothing. But the client tolerated the error, which enabled the author to go on to solve the problem.
Engineers must disturb normal unit operations simply to observe how the equipment responds to change. Errors are the price paid for progress.

PRACTICAL EXAMPLES

The troubleshooting work of Process Improvement Engineering provides many opportunities to solve process problems using common sense and the scientific method. Two cases will illustrate solutions to such problems.

The first case demonstrates how technical proficiency affects profitability in the short term. And the second case shows that solutions do not have to be complicated to be effective.

CASE 1

In 1980, a large U.S. Gulf Coast refinery concluded a 5-month strike. During the strike, the plant continued normal operations, manned by engineers and supervisors. At the conclusion of the strike, the accounting department reported:

Refinery weight balance had improved by 0.8% - Volumetric yield of liquid products had increased by 1.3% a Refinery profits had doubled.

By the year's end,,operations had returned to normal. Energy use was up, volumetric yield was down, and rotating equipment failures had increased by an order of magnitude.

CASE 2

A few months ago, an independent refinery hired Process Improvement Engineering to troubleshoot a problem. The heavy vacuum gas oil (HVGO) produced by the vacuum tower was black, rather than the translucent green color it should have been (Fig. 2).

A calculation of the vertical vapor velocity in the flash zone of the vacuum tower showed that resid entrainment was causing the black HVGO. The high vertical vapor velocity was a consequence of excessive light vacuum gas oil (LVGO) production. But most of the LVGO was diesel-oil boiling range material.

The refiner was aware of the poor diesel oil recovery in the upstream atmospheric crude tower. The crude tower flash zone pressure was reported to be 55 psig. The refiner's explanation for this high pressure was that the crude tower was flooding.

The unit technical service engineer concluded that flooding in the crude tower was causing a high pressure drop. The tower top pressure was 17 psig, hence the tower delta P was 38 psig. Further, new higher-capacity trays had been ordered to correct this problem, despite the fact that, other than the high flash-zone pressure, the tower was fractionating quite normally.

The unit engineer explained how she had calculated the 38 psig differential pressure from the two pressure recorders in the new, computerized control complex. When the pressure at the top of the tower ",as checked with a pressure gauge, however, the top pressure was found to be not 17 psig, but 51 psig.

The 17 psig pressure recorded in the control room was correct. But it was measured in the overhead vapor line 50 ft downstream of the column. The crude unit was shut down. The overhead vapor line was found to be 90% plugged. Apparently, the corrosion-control chemicals injected into this line had reacted with the olefinic slops being rerun on the crude unit and formed a rock-hard carbonaceous deposit.

When the unit was brought back on-line after clearing the vapor line, the flash zone pressure decreased from 55 psig to 29 psig. The new trays were left in the warehouse. Diesel oil production increased by 50%, and the HVGO color was a translucent green.

When the unit engineer was asked to check the LVGO product for diesel oil, she said that this was not possible because the LVGO production rate was now zero.

COMMON THREADS

The 12 principles outlined earlier are common, well-known tactics among most senior process engineers. Several common threads can be discerned among the principles:

They don't cost anything; therefore, unlike quality-in-excellence programs or advanced computer control, no one is selling them.
They assume motivated technical personnel, although this is not always the case.
They assume that an organization really wishes to break its dependence on meetings, although many do not.
They assume a technically competent staff, versed in the basic principles of engineering.

One ultimate tenet has arisen after having worked through more than 1,000 process problems and instructed as many process engineers in seminars: The key to enhanced refinery profitability is the application of basic engineering technology at the functioning level of the process equipment.