|Research has shown that new plant start-ups of dry particulate processes only hit 60 percent of their designed target rate. To address why this might be happening, Processing recently sat down for a conversation with Joe Zerbel, senior project manager for Hapman, and an expert on material handling applications. We discussed a variety of topics, including the unique challenges of particle processing applications, the need for a synergistic view of processing applications, common end-user pitfalls, application suitability, and installation best practices.
1. What are some of the unique challenges plants face in applications where the processing of particles is required?
The challenges plants encounter is numerous. The cause is often rooted in a failure to recognize the unique properties and behaviors that affect equipment performance. Even when some effort has been made to understand these behaviors, there is often a misunderstanding of how feed stocks from different vendors might vary or how the character of a material may change because of the process itself and exposure to heat, humidity or physical changes to the size and shape of the particle.
2. Knowing that the tendency is for technologists to work within their own specialties, how can processing plants promote a more synergistic view of the entire process?
Excellent question. It once seemed as though engineers who were responsible for a certain project were more familiar with all aspects of a given project’s requirements. Though they may well have been employed as a mechanical engineer, they commonly possessed a working knowledge of electrical controls, processes design, as well as operational, health and safety needs.
Today, for several reasons, inexperience and/or an overly narrow focus creates more opportunity for important details to be overlooked. Now, more than ever, there is a need for coordination across disciplines and interests to avoid costly mistakes. A common frustration among equipment builders is in finding that review of mechanical installation requirements and controls integration of their systems begins anew with new “customers” once it arrives on site. Then, once more, when the equipment is turned over to the plant operations people, another understanding of project objectives is revealed by a third set of customers. At each stage, as new expectations come to light, the prospect of achieving customer satisfaction is complicated by changing and sometimes competing interests.
Rightly, their focus is on the end result, even the result at each stage. They understand the quality measures of the product and aggregate rate requirements. But product quality indicators do not translate well to into terms that are helpful to equipment manufacturers and instantaneous processing rates at each stage can vary depending on whether they are performing discrete batch processes. Each stage of the process can change the material behavior, especially its flow or caking tendencies, as a consequence of changes in temperature, moisture and particle size – to name a few.
4. Can you provide an example of how one step in an industrial process affects the next step in the process?
Let’s say a plant uses lime as a process ingredient. They are a big consumer so for cost reasons they purchase pebble lime in bulk. The pebble lime is conveyed from a silo to a hammermill using a belt conveyor at a rate of 40,000 lbs/hour. They knew this worked because one of their engineers had conveyed pebble lime at his former job. So the plant installed two identical belt conveyors one to feed the hammermill and one to take the powdered lime away from it. To save time, they supplied a SDS sheet for the pebble lime to their belt conveyor vendor. It was a straight-forward application and they were in a hurry due to looming production deadlines. The vendor, in their sensitivity to the customer’s urgent situation and sincere desire to help did not press the matter further when her request for information was met by frustration from the overworked engineer.
So the vendor built and shipped two identical conveyors. The results were mixed. When the system started up, the first belt conveyor performed as expected delivering pebble lime to the hammermill at the prescribed rate. However, the first problem they discovered was that the take-away conveyor was overwhelmed, spilling material off from both sides of the belt. They came to realize that the hammermill did more than reduce the particle size, it also decreased the loose bulk density from 55 lbs/ft³ to 35 lbs/ft3, which meant a 57 percent increase in volume. To compensate the plant sped the conveyor up. The next thing the plant noticed is that the powdered lime stuck to the belt and a high percentage of it carried over, built up inside the frame conveyor frame and caused it to track poorly. So, they cut a hole in the conveyor to let the carryover spill out onto the floor and dedicated manpower to shoveling carryover onto the conveyor. The third thing they noticed was that quite a bit of lime dust became airborne, which was exacerbated by the necessary increase to the speed of the second belt. In the end, the plant was unhappy with the conveyor vendor because their conveyor didn’t perform well. Meanwhile, the conveyor company argued that they were not told that the second conveyor was handling powdered lime at a greater volume.
5. How can budget constraints have a negative impact on plant designs, particularly when it comes to pilot plant scenarios?
Budget constraints are nothing new. We must all find a way to work within our means. The biggest issue with regard to some equipment purchases is that plant managers can inadvertently convey the wrong message to their engineers and purchasers about budgets. While, most would agree that spending more on the equipment to save more on maintenance and operating costs can be justified, some justification procedures mire engineers in complexity. Part of the problem is that with perfect hindsight it is always easy to calculate the cost of a bad decision but weighing the increased cost of a piece of equipment against vs. theoretical operating and maintenance expense is not as so straight forward, quick or easy. It is not unusual for engineers to perceive less risk in cutting corners and working within budgets.
6. How can plants best troubleshoot adverse events and determine the likelihood that an adverse event may occur, as well as estimate the severity of each event?
For plants, troubleshooting upset conditions is always an easier task than predicting or preventing them, let alone trying to estimate the potential severity. The best plan is to partner with a competent, experienced vendor. Breadth of product line and length of service to the industry are indications of a vendor whose advice you can trust. A vendor that sells more than a single solution for a given need can more easily and more honestly discuss the upside and downside of each equipment, including all reliability factors, breakdown scenarios and the time required to mitigate them. It is worth noting that most are heavily dependent on the plant’s experience with their materials in various devices. When researching the application of equipment on a material that the plant has no experience with, it may be best to arrange for a test. Most vendors provide these.
7. Can you discuss the dynamic of substituting guesswork for data in industrial process applications and why that is a prominent and particularly troubling practice?
Guesswork is costly. It often begins innocently enough through a request for budget pricing. Plant engineers, lacking data will often ask vendors to “assume” certain conditions for the purpose of obtaining a quick budget price so they can arrange funding. The problem that can occur is that when the project receives funding, nobody goes back to verify whether the assumptions were correct. Occasionally, as a consequence of verifications, important differences are discovered having major cost ramifications or disqualifying the equipment altogether.
8. Why might a customer be unwilling (or unable) to provide sufficient details to allow the vendor to design equipment, and why is this a particularly troubling scenario?
Reasons differ for withholding information from vendors. There may be concerns about inadvertently revealing the plant’s intellectual property. Regardless of the reason, it is risky business. Equipment manufacturers typically have a sincere desire to help the plant avoid a costly mistake and can provide valuable expertise to positively influence the overall system design.
9. How has downsizing impacted the ability of processing plants to effectively select and specify process equipment?
Over the past several years the downsizing of engineering departments has lead to a knowledge drain, but I am sure there are a number of reasons why plants specify or purchase the wrong gear. It can be difficult for equipment manufacturers to know why, although some of the reasons have been touched upon previously. The one general observation is that such decision-makers seem to have in common an urgent need to do something and are thus impatient with questions.
10. How do particle processing applications differ from liquid or gas processing applications?
Fundamentally, from a handling standpoint, there are fewer variables to consider when handling material in gas or liquid form consisting mainly of holding vessels, pumps, pipes and valves, which naturally vary depending on viscosity, corrosiveness, etc. However, beginning the description of a particle, whether powder, granule, flake, fiber, pellet, extrudate, or prill, each has a specific behavior and flow character depending on numerous influential factors including moisture, temperature, particle size, particle shape. For those reasons, scholarly articles written on the subject have gone so far as to suggest that dry bulk particles exist as a fourth state of matter.
11. What role does process modeling play in particle processing applications?
Process modeling, in the way that I have seen it used, is helpful in determining the best way to replicate, upgrade or streamline existing systems by way of studying and documenting and analyzing efficiencies. With the ultimate goal of achieving greater reliability, while maintaining or improving quality and reducing operating/maintenance costs, engineers can identify aspects of a system which may be served better by different equipment, processes that are unnecessarily redundant or processes that can be combined.
12. Why is process modeling typically not a replacement for pilot plant operations?
For new products where actual experience is in short supply, there is a tendency to search out and rely upon empirical data for use in the development of a process model. While this may serve as a starting point, it should not be used exclusive of material tests or pilot plants and an awareness of the potential for various problematic material behaviors that can only be fully appreciated when exposed to the various handling and processing conditions. The cost of testing and piloting of processes pale in comparison to the cost of operating plants that are plagued by process inefficiencies, profit robbing upsets and high maintenance. Once the architecture is complete, the system is installed and promises have been made for finished product, it can be prohibitively expensive to circle back to address these.
13. What common pitfalls do engineers encounter when designing, operating and maintaining particle processing systems?
A properly designed system will inherently have the fewest number of operating and maintenance problems. Taking care to avoid the following common pitfalls will greatly enhance your opportunity for success.
Recognize that every material is different and behaves differently. Don’t assume that because it looks like something you’ve handled before that it will behave the same way. Make sure you understand how the particle size, temperature, humidity and blending with other ingredients will affect the flow character and any build-up tendencies.
Most engineers are careful to consider how new equipment may affect the plant environment, i.e., sound, temperature, vibration, corrosive vapors, etc. However, for some reason, it is less obvious to consider how the existing environment might affect the equipment. Make sure plant conditions are communicated to the provider of the new equipment.
Often the duty cycle is used to determine how robust a piece of equipment may need to be, opting, perhaps for a lesser machine, if the actual run time is low. However, in doing so, do not lose sight of the failure mode and what it involved in correcting the most likely occurrences. If failures are not easily predicted or rectified quickly enough, the money saved on the equipment might not compensate for the cost of even one unscheduled downtime event.
Required throughput rates are an easy calculation and rarely omitted from equipment specifications. However, there are often other performance requirements that are less obvious but just as important. If things such as particle attrition, segregation of blended materials, or packing of sifted materials are undesirable, make sure the specification so states.
It is not unusual for operations people to have so struggled with a particular issue with an existing piece of equipment that the specification for equipment to replace it makes that issue a priority. However, do not let tunnel vision obscure the other project requirements that may be just as important.
It is the goal of every equipment manufacturer to satisfy their customer’s needs. However, their ability to do so can be severely compromised if stakeholders are not on the same page prior to the equipment build. Too often manufacturers find themselves passed from one contact to another during the fulfillment of an order, discovering at each stage that the new stakeholder has different, sometimes competing requirements. Not only is it difficult for manufacturers to hit the moving target but in the end, overall system performance can be compromised.
14. What are the recommended best practices for designing and optimizing the performance of particle processing applications?
Aside from the advice already offered above, I suggest two things be kept in mind. First, there is no substitute for good information on the subject material and, generally speaking, there is no better source for material information than the consumer of the material. Actual experience with the material reveals much about its character and behavior that is not available from any empirical data source. For new materials and new processes, testing is never a bad idea. Most vendors provide such services. The second suggestion is to remember that with regard to interpreting the material information and selecting suitable equipment, the equipment manufacturers themselves can be the best resource. Especially those who have been around for a long time and offer multiple technology types for a given process need.
Application engineers for such companies can speak knowledgeably about the strengths, limitations and maintenance requirements of each, in a head-to-head comparison.