← 返回

2026.10.12 | TB | Metals | Kaiser Aluminum Warrick Ramp-Up

原标题：Kaiser Aluminum – Warrick Ramp-up & Capacity Outlook

#metals

Third Bridge · 2026-10-12

Analyst-led • 22 Sep 2026

Companies: KALU US, AA US

Andrew Wehr (AW) Owner & President at C&I Electronics LLC

Curren Larson (CL) — Third Bridge Sector Analyst

Agenda

Key Insights

Introduction

CL [00:00:02]: Welcome to Third Bridge's Interview entitled Kaiser Aluminum – Warrick Ramp-up & Capacity Outlook. I'm Curren Larson, and I will be facilitating today's Interview with Mr Andrew Wehr, former Capital Project Manager at Kaiser Aluminum.

Before we start today's Interview, please state I agree or I disagree to the following statement: You understand the definition of material non-public information and agree not to disclose any such information, or any other information which is confidential, during this Interview.

AW [00:00:32]: I agree.

CL [00:00:34]: Would you mind briefly introducing yourself before we jump in?

AW [00:00:39]: My name is Andrew Wehr. As previously stated, I'm a former Capital Projects Manager for Kaiser Aluminum.

Questions

00:00:49] Kaiser Aluminum's management has guided to holding Roll Coat 4 utilisation at roughly 80% to prioritise product quality and delivery performance. From your experience, what specifically prevents it from operating closer to its design speed today?

AW [00:01:24]: When I was still involved, the major roadblock was the oven and fume mitigation system.

CL [00:01:55]: Particularly regarding the operational aspect, what does that mean in terms of necessary capital investment or just a timeline to being able to clear that hurdle?

AW [00:02:13]: It means from a technical standpoint that a larger incinerator is needed to treat the vapours coming off the oven if you want to run the line faster from a practical standpoint, operational standpoint, that's an issue that has to get worked out contractually with the OEM, with the manufacturer. It's not an easy transition or an easy fix. It's a rather large component to that whole system.

CL [00:02:51]: What does it cost to implement a larger…

AW [00:03:03]: About one year ago, I probably would remember to tell you, it's well north of USD 5m to do something like that. Because of all the safety requirements, safety interlocks, there's a lot that has to be figured out to make changes. USD 5m might be a starting point.

CL [00:03:37]: To clarify, it's not something that can just be engineered out. Is this like a solution that strictly would have to be replaced?

AW [00:03:45]: It's not an easy solution for in-house engineering to just try to find a workaround. It's a physics question, a volume capacity question. At some point, a larger capacity of air transitions turnover is going to have to happen, and that means either bigger incinerator or bigger fans or something.

00:04:17] My understanding is that BPA [bisphenol A]-free coatings require tighter temperature control and slower line speeds. In practice, how much do BPA-free coatings reduce sustainable throughput vs the 1,000ft-per-minute design speed?

AW [00:04:34]: On the legacy lines, my understanding was that would reduce the line speed at 50%. ​There was a review of options like post-coating trimming where you would take the defects off the edges of the sheet after the coating was applied to try to help that, but I don't know if any of that was actually ever implemented. For most traditional coating lines, it's not an easy product to use and requires a slower pace on the line.

CL [00:05:30]: Those are for most coating lines – that's consistent across the product mix?

AW [00:05:36]: My understanding is it's challenging for everyone.

CL [00:05:46]: Is that something that you would expect to be meaningfully engineered around over time, or is that something that we should just understand is structurally going to be slower throughput?

AW [00:06:02]: No. I think given enough time between the coating supplier manufacturer and the companies that are actually doing the coating application like Kaiser, they'll work something out. That tends to be the trend or that's the trend that product lines are going to, so they'll have to work something out on that, whether it's a post-trim process or some other way of controlling product temperature or something, but I feel like it will eventually work itself out.

00:07:10] Roll Coat 4 can apparently cut changeovers from roughly one hour on the legacy lines to around 15 minutes. How meaningful is that reduction for effective capacity? I'm not familiar with how many product changes a facility typically makes.

AW [00:07:33]: This is a pretty significant upgrade really. Product changes kind of depend on just what the orders are, what orders they have. They try to limit the number of product changes as much as they can, but at some point, you've got to switch over to a different product. That was considered a game changer having that reduced timeframe because I think outside of, let's call it, unavoidable maintenance, that line change is the most downtime that you see. Cutting that to 25% is pretty ideal.

CL [00:08:17]: Could you give us an idea of the amount of effective capacity that would be conducive to? Is that essentially just higher utilisation?

AW [00:08:31]: Yes, higher utilisation. I wouldn't put a percentage on it, if they're 80%. I don't think I could say that makes it 85% or 90%, but it definitely makes up for a higher utilisation.

00:08:50] Once Roll Coat 4 reaches the point where it's operating consistently at commercial rates, what would you expect to be the next operational constraint to extracting additional output from the facility?

AW [00:09:11]: From my past experience, the next bottleneck was usually in the ingot department. Ingot is the start of the production process for all the coated lines, hot mill, cold mill and the coating lines. That seemed to be the area that was struggling the most. There were periodic issues where maybe the cold mill was more of a constraint, but some of those issues were addressed, but the ingot department seemed to be the biggest bottleneck downstream or upstream.

CL [00:09:46]: Are you familiar with what exactly is causing the bottleneck at the ingot?

AW [00:09:56]: To some extent, I think a lot of it had to do with experience retiring, new personnel coming in didn't have that level of experience on how to operate older equipment. I won't say that it was the condition of the equipment, not allowing them to produce what they need to produce. I think it was more of a lack of knowledge of how to produce ingots at the rate needed.

00:10:32] What limits the throughput at the hot and cold mills today? Is it maximum rolling speed, downtime or yield loss?

AW [00:10:58]: Majority of my experience was with Roll Coat 4, but from what I remember, they did have some issues as far as mainly just, I want to say, overall general reliability because a lot of that equipment was quite a bit older. There are some large VFDs drives that are very outdated. They're obsolete. Getting parts for those is very challenging, working on them is very challenging. There was even some foundation issues for equipment that had to be addressed at one point. It's really just age of equipment and increased maintenance that comes along with age of equipment.

CL [00:11:47]: As it stands today, what's the necessary investment? Would you expect a meaningful maintenance CAPEX step-up because of these ageing assets? Could you elaborate on that equipment bottleneck?

AW [00:12:07]: During my tenure, I think the next phase of capital investment was going to have to be in the cold mill and hot mill area to try to make them more reliable. Like I said, those drives, that's a very expensive replacement. That's a project that had been out there for a while that had gotten moved YoY to make room for other things, but at some point, that was going to come to the surface and have to be addressed.

00:12:48] In terms of Roll Coat 4 running below its potential, would you characterise the lost output as constraints within the coating line itself or more so the insufficient upstream side?

AW [00:13:17]: Specifically Roll Coat 4, my assumption would be that that would be a prioritised line because it is newer and a lot of customers are going to want to see product off of that line. I don't see that line being starved for metal. If anything, it's probably prioritised to get metal from cold mill, hot mill. If it's limited, it's limited by its current functionality that it's only able to run 80% or whatever of design.

00:13:50] To what extent does the age of the upstream equipment create meaningful quality or consistency issues for the metal feeding the coating lines? Is that a consideration, or does it go back to the coating line itself?

AW [00:14:10]: Ageing condition plays a big part. I mentioned previously, equipment foundations had to be repaired. There was an extensive outage that had to be taken to address that. While you're doing that one of your two cold mill lines is offline, there can be quality issues because of rolls, steel embodiment because of old rolls, that kind of thing. Yes. I think that is the main driver of issues. It's just the ageing current condition of the equipment.

00:14:50] If upstream equipment were to cause a quality issue, what would that look like operationally? Is it additional processing? Is it a matter of delayed shipments?

AW [00:15:17]: The way the system is supposed to work is that would create a high scrap rate. That material would go back to ingot, get remelted or reused. You would hope that if it was an off-spec material, it would not get through to coating. There's not really a way to rework without running it back through as scrap. It would mainly be a high scrap rate, which would starve your coating lines from production. Then ultimately, you would delay shipments because you didn't have product going out the door.

00:15:55] As of today, would you say the legacy ingot operation is reliable and up to standard to meet the requirements for Roll Coat 4?

AW [00:16:22]: What would it take to bring the legacy ingot operation up to reliability?

CL [00:16:28]: Yes, if you don't see it currently being…

AW [00:16:34]: I think what would help them is to ensure that they have a fully operational coated scrap melter. I know there's been a project out there for a while to bale side trim chips. When you run your aluminium coil through for coating, your defects are usually on your edges, so they'll trim a little bit of metal off of each side as before it goes through the line and those little trimmings can be brought back to scrap and reused in the process in ingot. The problem is all these light fluffy chips, they don't melt very well. They tend to burn. It's the same thing. Those melt better and you get a lot more recovery of metal. I think having a better way of recovering scrap metal, having a fully functional scrap melter would go a long way. I can't speak a whole lot to the condition of the furnaces. When I was there, I think they were in decent shape, but I know the scrap melter was always an issue of concern.

00:19:37] Kaiser is currently targeting USD 60m-80m of long-term sustaining CAPEX. Given the age of the upstream equipment at Warrick and the need for upgrades such as drives, how adequate do you think that level of spending is to maintain reliability and efficiency while still being able to leave room for growth capital?

AW [00:20:11]: I think that's a positive estimate. If I'm looking at this from a perspective of wanting to see them investing in that operation and improving those operations, I would be very pleased by that. I think that shows the commitment to upgrading those systems. USD 60m-80m, a little bit of it depends on the duration. Is that over 10 years? Is that over five years or whatever, but I think that's a very reasonable commitment to try to upgrade those systems.

00:20:45] If Kaiser decided to fully de-bottleneck the hot and cold mills today, what would the practical project timeline look like, considering both the engineering and equipment procurement aspects? How long of a project does that become, all the way through commissioning and reaching a stable run rate?

AW [00:21:21]: I think this is the one I flagged is difficult to answer because it really depends on the specific scope of the project. Is it just the drives? Is there I don't have a good idea of what current issues they have. I will say that some of the stuff we were looking at, you may have up to one year for procurement for large VFD drives. The install would be much quicker than that, but you're still probably looking at a couple of months of downtime to implement and probably one good year for engineering and procurement.

CL [00:22:00]: How are these decisions made in practice? When you think about the environment today with the current aluminium pricing, are you pushing these things off as far as you can? How should we think about the timing of which these investments are made?

AW [00:22:23]: Ideally, you're not pushing them off. I think it just comes down to the return and the risk analysis, really, are you at risk of losing your line for three months when it's not very inconvenient time, but you also obviously have to balance that with the operating income coming in so that you can afford it.

00:22:58] If Kaiser had one incremental dollar of CAPEX to deploy at Warrick, where do you think it would be able to generate the greatest incremental economic benefit today, considering all the aspects we've discussed, whether improving ingot production, upgrading the hot mill or further optimising Roll Coat 4?

AW [00:23:21]: I would have to say improving ingot production. I thought about this one quite a bit. To me, it's a trade-off to the ingot and the hot mill because with ingot, you can replace that operation by buying ingots directly. If that's the route you want to take, then I think the hot mill is your next target, but if you're not going to do that, then I think ingot is probably where you start.

CL [00:24:01]: Beyond the drives, legacy equipment and ingot inefficiencies that we've discussed, are there any less obvious capital requirements at Warrick that you think we should be paying attention to, or that you think are perhaps overlooked by investors?

AW [00:24:36]: I mentioned the chip baler earlier. It's not obviously necessary for production, but it is a way of increasing your recovery, your scrap recovery and making you overall more efficient. I think that's something that should be looked at.

00:25:25] As customer demand for low-carbon aluminium increases, how difficult is it practically for Warrick to source sufficient low-carbon primary aluminium if internal ingot production falls short? What alternatives exist outside of Canadian supply?

AW [00:26:10]: In the US, I don't know that there is an alternative outside of Canada, realistically. You only got a couple of smelters left in the US, and they're both very carbon-intensive sources. I think the only exception to that is a higher scrap incorporation. There's challenges to that, but apart from incorporating a higher level of scrap into your ingot development, I don't know that there is much of an alternative outside of Canada.

CL [00:26:48]: Given the trade tensions for Canada – Trump has announced he's trying to make a deal with Belarus for potash – what's the pragmatic solution? Is it just ensuring that you can have potentially more ingot production? If so, where does that come from? Can it be brownfield, or does it have to be greenfield? How should we think about that?

AW [00:27:31]: With current production techniques for aluminium smelting, it's your carbon inclusion for the standard process of making aluminium, your carbon reduction mainly comes from your power source on how you make that aluminium. Like in Canada, hydroelectric power is the main power source and is considered a very low-carbon energy source. In the US, it's either coal-fired or gas-fired, which is a higher carbon source. I don't think everyone is using carbon anodes to make aluminium. Nobody has got anything different. Apart from a US smelter repowering to a different low-carbon source, it'd be very difficult.

00:28:28] When Kaiser sources low-carbon aluminium from Canada, are packaging customers generally willing to absorb the resulting premium and tariff costs, or does the company ultimately bear some of that cost through lower conversion margins? Of course, this is dependent on what is done with the tariffs – whether they are reduced from 50% to 25%.

AW [00:29:00]: My understanding is it was always a pass-through and material cost was a pass-through to the customer, but what happens there is you just think about food packaging, what's the primary food packaging besides aluminium is plastic. If a food manufacturer has an alternative like plastic that's cheaper, then they don't have to buy as much aluminium packaging. What can happen is they can reduce their order down to maybe a contractual minimum. While that cost may be absorbed in the pass-through, the customers ordering less material. With oil prices now, obviously, that's an interesting conundrum for food packaging industry because both options are relatively expensive, but it just shakes out in less orders essentially.

00:29:58] Could you elaborate on how the changes in Section 232 tariffs on Canadian aluminium alter the economics of importing low-carbon primary metal vs investing in additional ingot capacity or using more recycled aluminium? What might be the economic impact, and what optionality would Kaiser have?

AW [00:30:37]: If tariff levels stay strong and if Kaiser can lock in a large quantity of scrap over time and keep a coated scrap melter healthy, they'll have an advantage with scrap, but I think that's a challenging thing to do. There's a lot of manufacturers going after scrap right now. Obviously, new coating lines have come online in the last year or two. There's a lot of people looking for scrap, but that's about the only alternative you've got to try to fight the tariffs is to find scrap to include in your process. In my mind, easing of tariffs is about the best option for them as far as lowering their costs.

00:31:30] When thinking about the entire aluminium rolling industry, what do you see as the biggest operational differences between a well-run facility and an average one? Is it maintenance practices, level of automation or equipment age?

AW [00:32:12]: Equipment age plays into it and economics kind of drive that to some extent. I think a big part of it is the people aspect. Management plays a big role in that, how management interacts with your frontline employees. I've seen in my experience, polar opposite of that. I've seen where that interaction is not good. There's not a lot of face time. There's not a lot of interaction with frontline employees, and you can have a prevalent attitude out there that if they don't care, why do I care? I've seen it the other way where management is very active. They ensure that employees know that they're engaged and they have concern over things. You see employees even from the same union, same structure, they tend to care more. They tend to try to squeeze what they can out of things. I think that goes a long way. That can only do so much. Obviously, at some point, the equipment is old and it's going to break. You've got to have a good capital plan. That's vital. You can't starve the system of capital for 10 years and then expect to get it past its life expectancy. That's unrealistic expectations. You've got to have a good capital plan. Leadership has to be on board with that plan, understand the ramifications if we don't execute this plan that we're not going to get X number of years out of our investments.

00:34:09] What do you think are the biggest differences between Alcoa and Kaiser, in terms of how they approach capital projects and maintain the assets?

AW [00:34:35]: There was a pretty big difference. Quite almost opposite. Alcoa has a very, very detailed system for managing capital. They have a very structured system for evaluating what you would call maintenance capital, evaluating the risk. A lot of different departments are involved in evaluating and determining that risk level. Even return-seeking projects, there's a lot of effort that goes into determining the return on those projects. There's a lot of verification that has to get done. Planning follows a very strict, what we call front-end loading or FEL process where you go through problem solving, early development, preliminary engineering and every phase has approval levels and deliverables and it's very structured, almost too much so, but it's very detailed and very structured.

You know the risks if you're not doing something. Kaiser is a smaller company. They're a little lighter on the overhead when it comes to management. While that gives them the ability to move faster on decisions, unfortunately, that also kind of made things easier on moving forward with projects before they were probably fully vetted or verified. That's one of the big issues we had with Roll Coat 4 was we probably moved a little too quickly in the beginning without knowing some of the inherent risk to that project.

00:36:42] Could you elaborate on customer qualification? What do the timelines typically look like when you add in new capacity?

AW [00:37:06]: I think if I remember right, that's somewhere around a 3-6-month process. It kind of depends on the customer specifically and what their coating product is. It depends on the variety of stuff you're sending that particular customer, but they're probably going to want to see the equipment run, see where you're at, see it is fully functional. You run some test coils, they work that through their system, they give you feedback. You run some more. It's just a buildup from there, but it's a 3-6-month process.

00:37:46] We mentioned the ageing assets at Warrick. What would you be most focused on, in terms of the telltale signs that replacement is becoming unavoidable?

AW [00:38:08]: I think the VFD drives are one of the biggest ones that have been sitting out there. Those need to be addressed. It's not just obsolete equipment. It's obsolete knowledge. The type of electrical engineering that went in developing those is no longer taught. You're reliant on consultants that retired 10 years ago to bring them back to understand relay logic. While new engineers can try to pick that up, they just weren't brought up in that, they weren't taught that and it's a risk needs to be addressed as soon as possible.

CL [00:39:06]: What's the downside indication of that? Would you start to see deterioration in utilisation?

AW [00:39:13]: I think you get to a point where you can't make repairs anymore, I mean you may be potentially at risk of taking one of the two cold mill lines off because you don't have enough components to fix it.

00:39:25] How do you expect Kaiser to balance low-carbon customer requirements, Canadian supply and tariffs, as well as internal production constraints over the next 1-3 years? What sourcing strategy is most likely to become more important?

AW [00:40:02]: My hope is that they're doing the best that they can to try to lock in sources for outside scrap, just whatever they can do for long-term procurement of scrap. They've got the coated scrap melter or it's kept healthy and it's able to melt coated scrap for them for use in their ingot department. They're also using their internal scrap wisely. I mentioned the chip baler. That would be a big help for recovery for their own material. On top of that, just working out production issues within the ingot department if they're going to continue to run ingot, getting those production rates back up to where they were 5-10 years ago.

00:40:51] Is there anything else regarding Kaiser that you think is perhaps not immediately obvious but we should better understand, whether production levels or something else, especially in the volatile pricing environment?

AW [00:41:23]: I don't know if there's anything completely apart. From my perspective, I look at it, we had at least two other coating lines coming online or come online. They're not just coating lines that have come on. They're hot mill, cold mill. Kaiser is competing against essentially new facilities even are upstream from where their coating line is. There needs to be a pretty tight focus on hot mill and cold mill efficiency in operations and reliability to stay competitive with the other players.

CL [00:42:05]: Then again, you had mentioned you think the USD 60m-80m maintenance CAPEX range more than covers that.

AW [00:42:14]: I was impressed to see that. I think to me, that means they're taking that seriously.

CL [00:42:19]: That's all my questions for today. I want to say thank you for joining me and sharing your insights into Kaiser and the broader industry.

AW [00:42:34]: No problem.

Transcription ends at 00:42:37 of the recorded material