Reasonable Assurance Dedication
By Greg Keller | Published in Nuclear News
I have been in the nuclear industry for 40 years, 32 as a supplier, with more than half of that with the Curtiss-Wright Nuclear Division. Several years ago, while participating on the team that updated the Electric Power Research Institute’s (EPRI) commercial grade dedication guidelines, I had an epiphany. It wasn’t until recently that I realized how this epiphany could lead to an industry solution. After discussing and refining the concept internally and with utility and industry leaders in the area of commercial grade dedication, Curtiss-Wright Nuclear Division is embarking on an effort with the potential to save the industry millions of dollars annually.
The epiphany occurred during one of the many meetings that ultimately resulted in EPRI 3002002982, Plant Engineering: Guideline for the Acceptance of Commercial-Grade Items in Nuclear Safety-Related Applications. My participation on the team was primarily to represent the suppliers of commodity-type products, such as fasteners, fuses, and bearings. The team had reached an impasse on what should have been a simple statement in the guidance document. I found myself in the minority, perhaps even the only dissenting voice in the room on the issue, but I was sure I was right. After what seemed like the entire morning had gone by with no real progress, it hit me—the argument we were having was based on a misconception.
It seemed that the non-suppliers in the room thought of the suppliers in tidy buckets: Third-party dedicators dedicated everything they supplied, and manufacturers manufactured everything they supplied. In reality, most suppliers are original equipment manufacturers, dedicators, and manufacturers. Some suppliers may fit into one category more so than others, but most fit multiple categories. The misconception in this case was that an Appendix B manufacturer manufactured everything it supplied. But while that Appendix B manufacturer was capable of manufacturing many products, the majority of the pieces supplied were in fact dedicated. Once this misconception was cleared up, the team was able to quickly agree on language and move on.
The current situation
The difficult market environment is having a major impact on the current supply chain. In an effort to remain or become cost competitive, nuclear plant operators are purchasing fewer parts and equipment, purchasing fewer units with each order, and waiting until the last possible moment to place a purchase order, all of which increase the suppliers’ cost to provide parts and equipment. The overall decrease in purchases increases product costs because overhead costs remain the same. Purchasing small quantities has an impact on product costs because one-time setup, testing, and other costs are spread over fewer pieces. And shorter lead times drive up costs when certain activities are expedited, overtime is paid, and other activities are interrupted. In other words, the steps being taken by nuclear plant operators to reduce total costs are causing an increase in unit costs.
There exists a large segment of parts and equipment in which testing and inspection costs are disproportionately high compared to the price of the commercial products. There are certain items, such as fasteners and fuses, for which the ratio of testing to product cost is so disproportionate that the product cost is essentially zero. These items are hypersensitive to purchase quantity, because when the $500 in testing costs is spread over two or 10 or 100 items, the unit cost for testing alone can be $250, $50, or $5 apiece, respectively.
In an effort to remain or become cost competitive, nuclear plant operators are purchasing fewer parts and equipment, purchasing fewer units with each order, and waiting until the last possible moment to place a purchase order, all of which increase the suppliers’ cost to provide parts and equipment.
Although the nuclear plant is purchasing hardware, it is essentially paying for testing and inspections. Based on this, the lot price for certain items might be $1,000 regardless of whether five, 10, or 100 pieces are purchased. This forces the buyer to choose between unattractive options. Let’s say a plant has an immediate need for five fasteners or fuses, and based on the small lot size and high testing costs, the supplier quotes $200 each for five pieces and $10 each for 100 pieces. Either way, the plant is paying $1,000. If the plant chooses to purchase 100 pieces, the items it actually uses are only $10 each, but it is placing $950 worth of items into inventory. And if it chooses to avoid putting unused items into inventory and purchase only five pieces, those pieces cost $200 each. Neither of these options is attractive, and if the plant chooses to purchase only five pieces and ends up needing one additional piece, it is facing another $1,000 purchase.
The suppliers understand these economics and could perform testing on a larger lot and place the additional items into their own inventory. There are several reasons suppliers don’t do this, but the primary reason is that suppliers do not want a large inventory. Suppliers recognize, as do plant owners, that inventory ties up capital and is a tax burden.
Make or buy?
Nuclear plants may purchase safety-related items from a manufacturer believing that the items purchased have been manufactured by that supplier. But in terms of individual pieces, the vast majority of items sold as safety related are, in fact, dedicated. This is due to simple manufacturing economics, whereby the cost to produce items is a function of material cost, setup costs, and run costs. Manufacturing in small quantities means the setup costs are spread over a small number of items. Imagine the cost to manufacture one smart phone. Not one of millions, but just one. The cost would be astronomical.
As described previously, smaller purchases lead to increased prices, and smaller purchases also drive a larger number of items to be dedicated versus manufactured as safety related. If it is hard to imagine a manufacturer capable of manufacturing an item in-house choosing to purchase that same item commercially and dedicate it, just look at the companies that make sealing products, fittings, and switches completely dropping their nuclear quality assurance programs, forcing the industry to dedicate commercial products as the only option.
In addition to purchasing smaller quantities of safety-related items, the industry is beginning to implement 10 CFR 50.69, which allows safety-related items that are categorized as low safety-significance to be purchased without imposing the requirements of 10 CFR 50 Appendix B and 10 CFR Part 21.
In addition to purchasing smaller quantities of safety-related items, the industry is beginning to implement 10 CFR 50.69, which allows safety-related items that are categorized as low safety-significance to be purchased without imposing the requirements of 10 CFR 50 Appendix B and 10 CFR Part 21. Instead, commercial grade items could be purchased, and alternative treatments could be applied to establish confidence in the items’ abilities to perform safety functions. If the current patterns and practices continue without a significant change in direction, the nuclear industry may find itself with a supply base incapable of providing certain critical safety-related products. What follows is a proposed way to change direction in a meaningful way, helping to reduce costs and maintain a healthy supply base.
The dedicating entity
One of the ways nuclear utilities are dealing with high prices for safety-related items is to dedicate certain products themselves. Rather than dedicate only those items unavailable as safety related in the marketplace, plants are dedicating items simply to save costs. There is a general belief that utilities are dedicating commercial items as a cost-saving alternative to items manufactured as basic components under a supplier’s 10 CFR 50 Appendix B quality assurance program. But for the vast majority of items, the choice is really between the plant dedicating the items or the supplier dedicating the items. Either way, the items are being dedicated.
Presumably, suppliers providing parts and equipment to the entire industry have greater purchasing power, more product knowledge, better test equipment, and more efficient processes than an individual plant, or even a larger fleet. So how is it possible that the suppliers are so much more expensive? There are certainly issues like profit, but the delta between supplier and in-house dedication costs is wider than can be justified by profits alone. It should be pointed out that if the nuclear industry’s suppliers were so highly profitable, they wouldn’t be leaving, or declaring bankruptcy.
The way many utilities purchase a large population of items from suppliers, the standard is full certification rather than reasonable assurance, and this has a huge impact on cost.
There is a larger issue at play here, one that the plants and fleets themselves may not be aware of despite being largely responsible for it. The way many utilities purchase a large population of items from suppliers, the standard is full certification rather than reasonable assurance, and this has a huge impact on cost.
Dedication v. certification
The underlying premise of commercial grade dedication is that performing sufficient tests, inspections, and other processes provides reasonable assurance that the part will perform its intended safety function. But many items, such as fasteners, fuses, bearings, and O-rings, are purchased generically, and no safety function is provided to the supplier. In the absence of a stated safety function from the purchaser, the supplier must treat all characteristics of the part as being critical and essentially verify all characteristics. In other words, the supplier is both dedicating and certifying.
A good example of certification-dedication is the purchase of safety-related generic fasteners to an ASTM material specification. The specification was written for general industrial use by organizations that assumed the manufacturer was producing 100,000-plus-piece lots, and so requiring $500 in testing is inconsequential to the product’s cost. But when a nuclear supplier dedicating a lot of 50 fasteners performs $500 in testing, each individual piece has $10 in testing costs alone. Assuming that the supplier in this case offers to sell these fasteners for $20 each, many nuclear utilities, faced with a unit price of $20 for something that can be bought commercially for a few cents, choose to dedicate their own fasteners.
The underlying premise of commercial grade dedication is that performing sufficient tests, inspections, and other processes provides reasonable assurance that the part will perform its intended safety function.
How can a utility dedicate fasteners for less money than a fastener supplier? The advantage the utility has is that it knows or controls where those fasteners go in the plant and can declare that certain characteristics are not critical, allowing it to perform fewer tests and inspections than the supplier. The utility may perform only $50 or $100 in tests and inspections on that same lot of fasteners and achieve a reasonable level of assurance that the parts will perform their safety function.
Suppliers may view this as an uneven playing field, but from a technical perspective, they are likely to agree that the subset of tests and inspections performed by the utility were, in fact, sufficient to achieve reasonable assurance. Perhaps the supplier may have chosen a slightly different set of tests and inspections than the utility did, and it is likely that the multiple utilities dedicating the fasteners each performed a slightly different set of tests and inspections. But regardless of whether or not there is 100 percent agreement on the exact number of tests and inspections and on which tests and inspections are performed, there is likely universal agreement that certification-dedication is inefficient and goes far beyond reasonable assurance. Plus, given that the individual site or fleet likely dedicates smaller quantities than a supplier that provides products to the entire industry, the choice is between two inefficient options.
The solution
Given that the utility is satisfied with the level of assurance achieved when dedicating items in-house and performing only a subset of the total tests and inspections required by the specifications, why doesn’t the utility purchase those same items from a supplier in a manner that allows the supplier to also perform a subset of the required tests and inspections? Part of the reason is a general lack of knowledge and understanding of what suppliers actually do.
For example, the utility buying a fastener from a fastener “manufacturer” likely believes that each item purchased is being manufactured as safety related. The utility believes it is saving money by dedicating commercial fasteners because fasteners manufactured as safety related are too expensive. But in most cases, this is simply a matter of two parties dedicating the same products to two different sets of standards—reasonable assurance versus full compliance to the specification.
But if the utility truly believes that performing five out of 10 tests and inspections provides reasonable assurance, why can’t it simply purchase the item that way from the supplier? The challenge is in how to communicate that. The utility either needs to send the supplier a dedication plan or it needs to provide a composite of all the safety functions for all the locations where the item may be used. The first is an administrative and perhaps also a policy challenge, and the second is a herculean task. And if multiple utilities sent suppliers multiple dedication plans that are all slightly different, the industry has still failed to achieve full efficiency.
Perhaps the best solution is for the suppliers of products to develop the dedication plans for the products they supply.
Perhaps the best solution is for the suppliers of products to develop the dedication plans for the products they supply. The suppliers know the costs associated with each test and inspection, and they know which tests and inspections are likely to cause a lot to be rejected. And frankly, the suppliers of products tend to be experts in those products. In addition, a fastener supplier that recommends forgoing an expensive machined coupon tensile test for a less expensive full-size tensile test likely has many years of test data to prove that fasteners never pass one test and fail the other.
When a utility selects a subset of tests to perform, it is likely that the availability of test equipment influences its decision. The supplier currently performing all of the required tests and inspections is in a better position to select the specific set of tests and inspections that provides the greatest level of assurance for the least cost. And to be clear, the main objective is not to minimize the cost, but to select the ideal set of tests and inspections that balances costs and level of assurance.
The dedication plan developed by the supplier is then provided to the utility for review and approval. It is anticipated that the supplier’s plan meets or likely exceeds the utility’s current dedication plan. The utility then adopts the supplier’s plan as its own, and purchases the items from the supplier (or any supplier) to the utility dedication plan. The utility gets the same or higher level of assurance as when it dedicates items itself, and the cost to buy dedicated products from the supplier is greatly reduced. The ultimate goal is for these supplier-provided dedication plans to gain wide acceptance and become industry dedication plans.
This basic solution is simple, but given the wide variety of products, as well as the multiple utilities and suppliers, the industry must undertake a coordinated effort to achieve all of the benefits and gain the greatest levels of efficiency.
Coordinated industry effort
Curtiss-Wright Nuclear Division’s Nova (a manufacturer/supplier of fasteners to the nuclear industry) could certainly develop and provide a set of reasonable assurance dedication plans to a site or fleet and begin supplying fasteners at a reduced price to that particular customer. But if another fastener supplier developed a similar set of documents and offered those to a different fleet, the industry would not achieve the full benefits of this concept. And one substantial benefit is to get a large number of sites purchasing the same items the same way, enabling all suppliers to dedicate in larger lot sizes.
While Curtiss-Wright Nuclear Division developed this concept and is working to create an initial set of reasonable assurance dedication plans, it recognizes that this needs to be an industry-wide effort for three main reasons. The first is that if the dedication plans are not shared with other suppliers, those other suppliers will create their own unique versions, and this concept will fail to meet all of the objectives. The second is that the Curtiss-Wright Nuclear Division does not supply every single product that can be dedicated, and the industry can benefit from using this concept for other product categories. The third reason is simply that the supplier base as a whole contains many product experts, and we need to tap into the entirety of the suppliers’ expertise, not just the experts within the Curtiss-Wright Nuclear Division.
Proof of concept
Fasteners are an easy first choice for a proof-of-concept effort. They are simple metallic items, manufactured to national standards, familiar to most people, and used throughout the plants in all types of equipment, including mechanical and electrical. Fasteners are available as safety related from suppliers and are also dedicated by utilities. And they fit this concept well because of a very important factor: The testing costs are disproportionately high compared to the commercial product cost.
With bolts, for example, the very expensive machined coupon test is not performed in favor of a less expensive full-size wedge tensile test. The tests themselves take approximately the same amount of time, but it is the preparation for the machined coupon test that actually takes the most time (machining the coupon). Destructive tests such as this are typically performed on a small number of samples, but certain non-destructive tests that might be performed on a small relay are typically performed on a larger sample, perhaps even the entire lot. The ability to forego certain tests on a relay not only saves a considerable amount of time by not having to perform certain tests, but also by not having to set up for these tests. An initial look at one particular small relay indicates that roughly half the cost to dedicate a lot size of 10 could be eliminated, while still achieving a reasonable level of assurance.
Not all dedicated products have an extreme ratio of testing to product cost, but enough do that if this effort is successful with fasteners, the plan is to expand it to other items.
Not all dedicated products have an extreme ratio of testing to product cost, but enough do that if this effort is successful with fasteners, the plan is to expand it to other items. The Curtiss-Wright Nuclear Division is committed to guiding this process to the point where it either launches successfully or fails to gain traction. But when we reach the point where it launches successfully, other suppliers will be invited to participate, and Curtiss-Wright Nuclear Division will back away from coordinating this effort and simply become one of many participants.
Developing the process
Curtiss-Wright Nuclear Division has discussed this concept with members of EPRI’s Joint Utility Task Group (JUTG) steering committee, as ultimately it makes the most sense for JUTG to coordinate this overall process, especially since it developed the guidance document for commercial grade dedication that was endorsed by the Nuclear Regulatory Commission in Regulatory Guide 1.164, Dedication of Commercial-Grade Items for Use in Nuclear Power Plants. Several utilities have stepped up and agreed to review the dedication plans developed by Curtiss-Wright, provide feedback, and help guide the initial process.
The early development phase of this effort will not just be to create reasonable assurance dedication plans for fasteners, fuses, and other products, but to develop a process and guidance on how to apply the lessons learned through the creation of plans for simple products so those lessons can be applied to more complex products.
One concept being considered is tightening certain requirements to compensate for tests and inspections that will not be performed. For example, Nova recommends performing only a full-size wedge tensile test for bolts and eliminating the machined coupon tensile test, which is the single most costly test due to the machining of the coupon. Given that the minimum tensile requirements for these tests are identical, perhaps the minimum threshold for the full-size test should be raised. If the minimum tensile requirement is 125,000 psi for both tests, and only the full-size test is performed using a minimum requirement of 127,500 psi, would this provide greater assurance that the machined coupon test would also have passed? If this concept is sound, is 2 percent the correct margin?
Perhaps the dedication plans developed by Nova will include tightened margins. Perhaps that margin will be as suggested above or will be based on a different percentage or some other set of criteria. Either way, the thought process used will be well documented so that it can be used by other product experts to develop similar dedication plans for different products.
By the time this article is published, there will be several dedication plans created and refined by Curtiss-Wright Nuclear Division and the participating utilities. While these parties have committed to launching this effort, in order to fully achieve the stated objectives, suppliers other than Curtiss-Wright Nuclear Division and a larger number of utilities need to be involved. Parties interested in participating in this effort should contact the author.
Conclusion
The last several years have been difficult for the nuclear industry, and change is needed in order to become or remain cost competitive. While many changes have occurred in the supply chain arena, most have been incremental. Utilities purchasing fewer items is change, but not innovative change. Similarly, sites and fleets choosing to dedicate products in-house is also not an innovative change, because all that is really different is the dedicating entity. An industry-wide effort to establish and implement standard dedication plans that permit the dedicating entity to achieve reasonable assurance and be cost-effective has the potential to save millions of dollars annually. It also allows utilities to focus on their core competency—the safe and reliable generation of emission-free electricity—and helps maintain a healthy and solid supplier base. The goal is to have reasonable assurance dedication plans completed for a variety of applicable items during the first half of 2019. The results of this effort will be presented at the EPRI Joint Utility Task Group’s Procurement Forum in August to a group composed largely of utility procurement engineers and suppliers that routinely engage in commercial grade dedication.
The last several years have been difficult for the nuclear industry, and change is needed in order to become or remain cost competitive.