Determining the best hardware quote

July 25, 2005
Keep a few pointers in mind when sorting through proposals for distillation column internals. They will make it easier to equitably compare the quotes, including assessing performance claims and even alternative hardware recommendations.

In our previous article, “Distill your quote request” (June, p. 30), we discussed how to prepare requests for quotation (RFQs) to get adequate and comparable quotes from tower internals vendors. It is common to obtain quotes from more than one supplier for price comparisons and checking each for correct device model names/numbers is easy with standard internals types. But if the project involves high capacity or proprietary devices, you will need to compare the offerings from the different vendors in terms of relative capacity, efficiency and pressure drop. You also must be prepared to assess any performance claims or recommendations from the vendors as well as suggested alternatives to the base case.

If your RFQ included a detailed base configuration, you probably have already checked that the requested device(s) will meet your process requirements. In this case, evaluation and comparison of the vendor base quotes is relatively straightforward. However, if the RFQ base configuration was unspecific or you are offered one or more alternative configurations, the first step in comparing the vendor quotes is verifying that the offered device(s) will meet your process needs.

Verify the device
When attempting to verify the suitability of an internal device in a particular distillation service, the best method by far is to have independent, unbiased tests of device capabilities. Such tests are available through organizations sponsored by industry such as Fractionation Research, Inc. (FRI) or academia such as the University of Texas’ Separations Research Program (SRP). Of course, your company must belong to an organization to be able to use its resources.

Lacking access to FRI or SRP, you still have some ways to check the applicability of proposed internals. However, these methods depend more heavily upon vendor representations and should not be considered independent assessments of vendor wares.

The first approach is to request actual plant data from a commercial application of the device. Keep a few critical points in mind: First, the tower service and conditions should as closely as possible match your proposed application and the tower configuration (number of trays or packed beds, tray or packing types and other internals items) should be provided. Second, get complete material balance information, including analyses of all feeds and products. Third, have the vendor supply unreconciled heat-balance information, including all redundant ways of measuring heat input and removal. You should plan on developing a simulation model of the tower to find the magnitude of the plant heat-balance error. This lets you determine the effect on the calculated device efficiency when the error is assigned alternatively to the overhead condenser/reflux and then the reboiler — that is, complete heat-balance data allow you to perform a sensitivity analysis.

Plant test data with complete heat-balance information often are very difficult to obtain. As a last resort, ask the vendor to share its in-house test results under a secrecy agreement. Usually these data are for air/water or air/isopar systems. (Isopar is a high-flash paraffin oil from ExxonMobil that has a surface tension of about 26 or 27 dyne/cm at ambient conditions, versus water’s 60 dyne/cm; air/isopar data suit hydrocarbon or other low surface tension applications.) Of course, air/water or air/isopar data only give capacity and pressure drop information, not separation efficiency. Exercise caution when interpreting vendor in-house data because results from only the most favorable runs might have been provided.

Also, talk to competing suppliers about known failures of a device. A rival vendor sometimes may offer such information to promote its own product. Once again however, you must be careful about the details because failures can occur for a number of causes not involving the device itself, such as improper design of feed locations or draws, or inadequate ancillaries such as condensers, reboilers or pumps.
 For packing applications, the vendor should offer the most efficient packing that meets the capacity requirements. Structured packing is preferred for sub-atmospheric pressure applications due to its low pressure drop, but should be avoided in hydrocarbon services above 100 psia due to erratic height equivalent to a theoretical plate (HETP) performance known as the hump effect.

Other process requirements
After verifying that the device(s) offered satisfy your capacity, efficiency and pressure-drop criteria, compare the base offering with other requirements set forth in the RFQ to confirm that all important criteria have been met. Considerations may include:

• process flexibility (turndown);
• mechanical strength;
• material type and thickness; and
• fouling resistance.

For packing applications involving a gravity distributor, a distributor test plan and cost should be included.

If you have alternative proposals in addition to the base proposal, the above checks should show if an alternative is better than the base proposal in ways such as fouling resistance, capacity, efficiency and/or pressure drop. They also should identify the process advantages necessary to justify additional spending for costlier alternatives.

Scope of supply. This should be checked carefully, especially when comparing quotes from different vendors. Gaps generally reflect details that may not have been spelled out or were simply overlooked. Here are several commonly missed items:

• new internal piping from existing nozzles, where necessary;
• gaskets for all pipe flanges (and correct gasket materials);
• undesirable “plate flanges” in pressure-containing piping, such as spray headers;
• differences in materials between vendors, especially for tower attachments and bolting; and
• structured-packing sheet thickness.

Structured-packing thickness is a major cost factor but also can be very important to constructability. To trim bid prices, vendors sometimes quote material as thin as 0.005 or 0.004 inches. This can be an issue for larger columns where workers go inside during installation. Crushing of such thin packing can be disastrous to performance. Yet, installers may cover crushed packing before it is detected. So, the risk of crushing must be recognized and addressed in the work process. Controls include time-consuming crimp straightening measures or requirements that workers stand on plywood sheets.  For critical path work, the added installation time due to such controls may cost more than the price of thicker packing material. Some operating companies do not accept sheet thicknesses less than 0.006 inches if installers will work on top of the packing, and some require even thicker material for high-capacity (large crimp) structured packing.

Exceptions and clarifications. If you included a set of engineering specifications in your RFQ, a vendor may take exception in its bid to some of the requirements. Be sure to determine the acceptability of all exceptions and request bid addenda, as necessary, to comply with non-negotiable specifications.

Ascertain exactly what equipment is being offered. For example, a vendor may list a support grid for a packed bed but may depict several types of such grids in its literature. Do not assume that any particular item or model of equipment will be provided unless it is specifically spelled out in the bid text or scope of supply. If there are any questions at all, ask for clarifications!

Process guarantees. If such guarantees are necessary, be sure they are included in the bid along with acceptance criteria that details what must be done or shown to activate each guarantee.

It is not uncommon to award different towers in a multi-tower revamp to different vendors. Occasionally, even different sections of a particular tower go to different suppliers. In such circumstances and others, vendors may not offer guarantees. This may make project managers uncomfortable, but is there much real benefit to process guarantees?

Consider the value of a process guarantee for a given distillation tower. If one or more separation objectives are not met, the vendor generally must supply a different internals configuration to fix the problem. Installation costs may or may not be included, but consequential damages are nearly always excluded from process guarantees. This latter point is key: economic losses from lost production, especially from a second unplanned shutdown, generally total many times the value of a process guarantee. In addition, there may be a period where the vendor attempts to fend off process guarantee claims, citing deficiencies or inconsistencies in operating data, flaws in the design simulation, feedstock differences from design, under-designed tower ancillaries, improper internals installation or even damage from operator error. Simply demonstrating that internals are the root cause of the problem may be a daunting task, with the plant losing money all the while. Considering all of the facets of this situation, process guarantees on tower internals can be nearly worthless to the plant owner. The real objective of the process engineer is to determine that the chosen internals will work so there is no need to resort to process guarantees.

Schedule issues. When negotiating delivery schedules, it often is useful to ask the vendor for an assessment of its projected shop loads. Because tower internal revamps require a shutdown, some project managers give a fictitious shutdown date of about a month earlier than actually planned to ensure that the vendor does not bump the job for other, possibly higher margin orders.

If on-time delivery is crucial, as it can be, for example, in projects where tower work is on the critical path, you may wish to negotiate incentive/penalty clauses on approval drawing receipt in addition to finished internals shipping. Most vendors state that the main bottleneck in the design/approval/procurement/fabrication cycle occurs in the engineering design phase. By adding such clauses to the approval drawing process, project teams usually can expedite this limiting step.

Make sure the proposed time allotted for client approval of drawings is reasonable. For simple orders such as a single new tray design, a day or two may suffice to review drawings, transmit comments and have changes incorporated. For complex, multi-tower revamps, at least two weeks should be provided for the return of comments; there also should be allowance for follow-up review of modified drawings without impacting shipping schedules. This should be negotiated during the bid phase.

It can be useful to assess a vendor’s responsiveness and willingness to accommodate changes or unusual requirements. However, initial impressions can be misleading because internals vendors often have separate sales and project execution staffs; implied cooperativeness may not survive the handoff between these groups. One tactic is to ask the vendor to name the lead execution person or project manager in the bid and start a dialog with that person before the bid is accepted.

Material shortages and surcharges. Steel now is in short supply, particularly in the United States. This has caused uncertainty in both the price of raw materials and their availability for fabrication. Some internals vendors have responded with materials surcharge clauses in their bids, with the contract price subject to adjustment based on steel cost at the time of order. Other vendors may still offer firm quotes, especially for small orders for which sufficient inventory of raw materials is in stock. In cases where a surcharge clause has been included in a bid, check to see if any competing bid offers a firm price. If other factors make it desirable to pursue a bid that contains a surcharge clause, try to negotiate with the vendor to have all raw materials purchased upon acceptance of the bid (or receipt of the first payment) rather than letting surcharges accumulate on an ongoing basis.

John G. Kunesh is a part-time consultant on distillation, based in Red River, New Mexico. He served as technical director of Fractionation Research, Inc., Stillwater, Okla., until his recent retirement. E-mail him at [email protected].

Raymond M. Sowiak is a senior process engineering specialist at Sunoco, Inc.,  Philadelphia. He is Sunoco’s technical representative to FRI and is a member of FRI’s Design Practices Committee. E-mail him at [email protected].