Engineering: By Owner or Supplier?

Stephen M. Hall, PE

Can generalists successfully deliver new process systems for their chemical manufacturing employer? The answer is a qualified “yes.” Engineering can be pushed to suppliers, as long as the Owner maintains perspective and pays attention to critical aspects of the design and installation. Generalists do have the training and skills needed to ensure that new systems are fit for their intended purpose. But their success is affected by their work environment and the support provided by their employer.

Chemical engineers are involved with every phase in the life cycles of process systems. The engineers are employed by Owners, engineering consultants, equipment resellers or agents, equipment manufacturers, construction companies, and maintenance contractors. Cost efficiency – and staffing streamlining – is on the mind of most operating companies. The in-house engineering talent is called upon to fill many roles. Consequently, people with broad, general skill sets are often preferred to those with in-depth knowledge about specific technologies.

Engineers working for Owners or contractors frequently delegate detailed engineering calculations to specialists who work for equipment suppliers. The suppliers are intimately familiar with their equipment, they possess the necessary design tools, and they have exposure to a wide range of industrial applications. But their expertise cannot replace the Owner’s responsibility to clearly and fully define the problem being solved. Delegation of sizing and selection engineering must be accompanied by reasoned and attainable specifications. The Owner must also provide proper context for the equipment including assurance that the specified equipment is appropriate for the function it will be used for.

Engineers representing the Owner – often a plant engineer or engineering consultant – are usually generalists responsible for managing all aspects of their projects. They often write performance specifications and rely on the suppliers to provide fully engineered instruments, equipment, and modular systems. This is where their requirements should be fully articulated so the supplier receives all of the information needed to design the equipment. Performance specifications are often under-engineered; this leads to disappointments and costly errors.

This article presents three real life examples where the Owner could have saved time and money if they had spent more on their own engineering before engaging with their suppliers. The examples are taken from my recent experience, when I’ve had an opportunity to observe the behaviors of Owners and suppliers.

Case 1: Heat Exchanger Design

Suppliers are proficient at doing the thermal and mechanical design of heat exchangers. They use specialized software that calculates heat transfer and pressure drop, given configuration parameters like tube diameter and number of passes. Those parameters can be adjusted in a seemingly endless number of combinations; the designer is responsible for selecting the combination that satisfies the heat transfer requirements while also meeting other project needs. If the Owner fails to fully understand and communicate the project’s objectives and constraints, the supplier may not provide an optimal design.

Situation: The Owner specified the thermal performance requirement appropriately, but did not provide physical properties for the process stream, a syrup with properties discoverable on the internet. The only configuration requirement given was that the exchanger utilize U-tubes with a single pass (technically, this is a two-pass exchanger). Multiple proposals were solicited and the low bid was accepted.

Result: The supplier found viscosity data for the syrup at the average duty temperature. They clearly indicated on the data sheet that the Owner did not provide properties and that they utilized the published values. However, in doing so they also assumed that the viscosity did not change with temperature, and did not recognize that the syrup would crystallize if its temperature approached to within 20° of the coolant temperature. After the exchanger was installed, but before being commissioned, a peer review revealed these shortcomings in the supplier’s understanding. Because of the high viscosity of the syrup, the temperature at the tube wall would be well below the crystallization point, and the exchanger was doomed to fail.

Remedy: The process flowsheet had to be reconsidered, and a different approach implemented for cooling the syrup. The problem was caught early enough to avoid a schedule delay, but there was a very significant cost penalty.

Recommendations: In this case, the Owner’s engineer would have recognized the issue if they had spent time to collect and analyze the process data. Heat exchanger data sheets have input fields for physical properties evaluated at the inlet and outlet temperature of the exchanger. It is also well known that the temperature at the tube wall is close to the temperature of the fluid with the highest film coefficient. With a high viscosity syrup on one side and a heat transfer fluid on the other, it is obvious that the wall temperature would be near the coolant temperature. Take the time, and exercise the care, needed to fully complete heat exchanger data sheets. Include all of the mechanical aspects such as tube size, overall length, number of passes, etc. Many configuration decisions can be left to the supplier’s expert designer; but the Owner should use the data sheet as a checklist and think about each of the choices to determine if there are any that must be constrained (e.g., length) or directed (e.g., number of passes) due to process or project requirements.

Case 2: Stirred Tank

Large manufacturing companies and engineering contractors used to have vessel engineers on staff. This is rare today, generally limited to groups who are engineering high pressure vessels in, for example, lethal or nuclear service. For standard tanks and reactors, the normal practice is to rely on vessel manufacturers to provide all of the calculations for determining wall thicknesses, nozzle reinforcements, and other details of the vessel design. This does not relieve Owners from their responsibility to specify materials, dimensional constraints, and process requirements.

Situation: The Owner specified the “working” volume of a vertical stirred tank, the minimum batch volume to be mixed, the maximum tank diameter, maximum overall height, type of heads on the tank and minimum clearance beneath the bottom outlet nozzle.

Result: The selected supplier took exception to the dimensional constraints because the specified volume tank could not fit within the diameter and height envelope, so a compromise was reached. The supplier agreed to the minimum batch size requirement and the tank and agitator were fabricated accordingly. After the tank was installed, the Owner discovered that the mixer could indeed function at the specified minimum volume, but this didn’t meet the process requirement because the batch instructions called for an initial charge that was one-third the volume of the final batch. It was the initial charge that needed to be mixed; as a result, the actual minimum batch size that could be accommodated in the vessel filled the tank to the top tangent. This was the “working” volume that the tank supplier designed to, but the Owner had assumed that “working” volume would leave headroom to allow for mixing turbulence.

Remedy: To utilize the equipment as purchased, the Owner was forced to revise their production plans and make full tank batches every time. They also had to monitor the agitator speed, and reduce it when the full tank volume was approached so that material would not splash into the head space.

Recommendations: Owners should use multi-discipline teams to walk through the operation of equipment and systems. In this case, the minimum batch size problem would have been recognized because the operations team would have called for starting the mixer after the initial charge and the engineers would have realized that the impeller wouldn’t be covered. Basic calculations should be performed for all aspects of the design – even if they rely on rules of thumb rather than state-of-the-art detailed algorithms – to confirm that specifications and designs obtained from suppliers make sense.

Case 3: Factory-Assembled System (“Skid”)

Small process units, usually fitting on one or steel frames and sized for transport by truck, are designed, assembled, and tested at the supplier’s shop. This approach is used for pump skids, clean-in-place units, ejectors, chillers, reactors, strippers, and numerous other process units. There are many attractions, and factory-assembled systems are very common. But the Owner needs to pay close attention to detail to obtain the most satisfaction. These units are often designed and fabricated to the Owner’s specifications, and if those specifications are vague then the results are unpredictable.

Situation: The Owner developed a process in the pilot plant that involved mixing two process streams together and immediately homogenizing them. For the scaled-up commercial unit, they decided to specify a factory-assembled module. Performance requirements were provided to the selected supplier, including flow ratio control accuracy, flow turndown range, temperature control, and the total permissible hold-up volume in the skid’s piping.

Result: The selected supplier was experienced at designing and building process units, but they used mechanical and electrical engineers for the work. They did not have the knowledge to evaluate or appreciate the process constraints that the Owner had specified. As a result, they agreed to the project not knowing that the specified parameters were impossible to achieve with the proposed design. A peer review recognized the problem and the project was temporarily halted.

Remedy: The Owner engaged a consulting chemical engineer to fully engineer the P&ID, including selection of the critical components, to ensure that the pumps, control valves, and other components would work together to meet the process requirements. During this effort the Owner had to relax the turndown specification because it was incompatible with other more important needs.

Recommendations: Owners should pay the same attention to factory-assembled units as to field erected systems. The P&IDs for factory-assembled units are critical and they should be held to the same standard as those for the stick-built units. Use Good Engineering Practices for specifying, selecting and checking the components used in the system to provide maximum assurance that the skids will meet the performance requirements.

Summing Up

Equipment suppliers employ engineers who possess – or learn – the specific knowledge needed to successfully design and deploy their products. But they cannot know the environment in which their equipment must function. It is the Owner’s responsibility to understand the application and ensure that the equipment is properly specified.

The three case histories presented in this article report instances where Owners relied too heavily on their equipment suppliers. This resulted in rework and unnecessary cost, as well as potential time delays. The situations could have been avoided ensuring that the Owner’s scope (or the scope of his engineering contractor) included sufficient design work (equipment evaluations, calculations, flowsheet creation) to fully articulate the specifications.

August, 2013