← 返回

2026.09.24 | TB | Energy | Marathon Petroleum Margins

原标题：Marathon Petroleum – Margin Sustainability, Utilisation Outlook & Competitive Positioning

#energy

Third Bridge · 2026-09-24

Analyst-led • 10 Sep 2026

Companies: MPC US

Michael Barham (MB) President at TooTall Engineering LLC

Curren Larson (CL) — Third Bridge Sector Analyst

Agenda

Key Insights

Introduction

CL [00:00:01]: Welcome to Third Bridge's Interview entitled Marathon Petroleum – Margin Sustainability, Utilisation Outlook & Competitive Positioning. I'm Curren Larson, and I'll be facilitating today's Interview with Mr Michael Barham, former Principal Engineer at Marathon Petroleum.

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.

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

CL [00:00:33]: Could you please briefly introduce yourself before we jump into the Interview?

MB [00:00:38]: My educational background is chemical engineering. The company I've worked for is Dow Chemical. Of course, Marathon's last one. I was with a company called Tesoro that Marathon purchased about 10 years ago, and before that, I was with Chevron, where I've been process engineer, controls engineer, controls manager at a refinery and worked for Philips. Well, now it's Phillips 66, it was ConocoPhillips. I also ran continuous improvement programmes, refineries, and corporate, Six Sigma, that kind of thing. In Chevron, I was emerging technologies adviser across the corporation, upstream, midstream, downstream. Then at Tesoro, I was a senior executive for controlled operational technologies and cyber and a few other things. Then at Marathon, I became principal engineer at Marathon.

Questions

[00:01:43] How would you characterise today's operational flexibility and constraints of Marathon Petroleum's Gulf Coast refining system, specifically for Garyville and Galveston Bay? Where do you see the biggest limitations for running these assets harder or adapting the crude and product slate?

MB [00:02:23]: I'll start running them harder. Actually, it's a low risk for them to run them hard because their history with both of those refineries, they've pushed them hard, past nameplate, which you can do that and you get a new design rate and for operations. They've put a lot into it, but traditionally, each one is a little bit different. Garyville, Marathon has owned it just about… Well, they owned it before it was born. When it was almost finished built, they bought it and then they completed it. Their design of it, it's a linear refinery. It's really phenomenal that way. It's very predictable that way. Of course, the history of the Texas City refinery is quite well known, but since that explosion and everything has been built up. Marathon has put an enormous amount into it, making it the largest refinery in North and South America. The sense I want to give with that is not only is it large, like Garyville is very large, very linear, very predictable, but Galveston Bay is a little bit more special because it can handle just about anything.

The variety of things it can handle and it can shift and move. They've handled the complexity of a site like that very well and to where it can shift and move what crude slate is opportunistic, making it how profitable it is right now. The second part of that, you asked about the crude slate. It's interesting with that because now you can expect there's going to be a lot more Venezuelan crude coming in and especially these two refineries. They do currently get a lot of Canadian select and which has its own issues, but with the Venezuelans, Galveston Bay has a little bit of an edge on being able to process that because that stuff is high in metals and they can handle, they have a lot of hydrocrackers and coking units to handle that stuff. It can hurt the cat crackers quite a bit on the catalyst and anything that uses catalyst, but they plan for that by increasing their sales and get better contracts with their zeolite providers. I think it's a very positive outlook for Marathon, especially with the Venezuelan select coming in.

[00:05:14] How should we think about turnaround schedules? If either one was meant to be seeing a major turnaround in autumn 2026 – if that were to be deferred into Q1 2027, how should we think about the implied risk of doing so, especially considering the higher content of western Canadian and Venezuelan crude?

MB [00:05:47]: Yes, and that's not unheard of. Sometimes you have to shift those for a few months, and usually, you have to shift them because some cat cracker or something has got some problems. There's a little bit of a limping until you get to the turnaround, and that's a good cost avoidance because if you do it early, it's just a mess. It costs a lot, just penalties for the turnaround crew you're hiring. I would just say, coming from different companies, Marathon is absolutely phenomenal at their turnarounds. They're usually on budget or under budget, and they're on schedule. That being said, these two refineries have some vulnerabilities with that. Most of it is weather.

It's great to have them in the fall or even later in the fall into January is not bad because you're past the hurricane and adverse weather conditions. Now, what hurt them a couple of years ago was that strange freeze that came out in south Texas, which you can't plan for, so abnormal conditions can affect it, but that's unpredictable. As far as their prediction basis, if they have to move into a few months, it's really not a major problem that the risk actually comes in with Venezuelan crude onto the catalyst, and you may have to lessen the cat cracker a little bit, which means you're making a little bit less gasoline-grade material, and being that type of season, that's not a major problem. It'd be a major problem in June, but it's not as major of a problem if that's the only constraint that's hitting you.

CL [00:07:34]: Would you expect Marathon to push a typical September or October 2026 turnaround into 2027, given where the crack spreads are, or do you see some economic degradation in doing so and therefore the company is better off just accepting the turnaround?

MB [00:07:50]: There really is an economic degradation in changing that schedule. You add in, not only some economic risk because you're hiring an enormous amount of contracting elements to come in and their availability is based on the schedule you already have. If they have to change that, then there's a substantial financial penalty per contract plus and you're not going to get the expected personnel. They'll fill personnel, but you're not really sure who you're going to get. The only reason for that to shift is not solely economic. It's mostly there's some kind of material problem or weather issue or something that is a little bit out of ordinary.

CL [00:08:41]: You mentioned that there is a financial penalty for changing the schedules. Could you elaborate on that?

MB [00:08:47]: Yes, your supply, your contracting elements, if you've got company X and company Z to come in and do turnarounds and they have the personnel and they do a whole variety of jobs, but those guys are contractors, so if they're not working on site, they're losing money. Their schedule is always full. If you try to advance the schedule, they have to take away from a current customer somewhere else, and you'll have to pay quadruple their losses or something like that. Of course, they will have to pay some smaller company saying, "Hey, we've got to defer," and then it's a penalty on them. It just stacks up like Russian nesting dolls on the penalties.

[00:09:42] Thinking about the cheap WCS [Western Canadian Select] in conjunction with the Venezuelan grades coming back into the market, there is some uncertainty about when those volumes really show up in scale. From an engineering perspective, what constraints at Garyville and Galveston Bay effectively determine how much heavy crude Marathon can process economically? How do the constraints differ when you think about WCS vs Venezuelan grades, given that management said Canadian heavy wins on economics roughly 80% of the time?

MB [00:10:31]: These constraints are actually, they're a little bit different each one. Now the benefit they have is both of them have a lot of the coking units, so the coker drums and all that, and you have large cokers, multi-drum cokers and hydrotreaters. Therein is also the constraint and the problem because they have to have a great cycle time. Now the constraint on what crude selection they're going to get is going to call the carbon content, the Conradson carbon content, which Venezuela is just a little bit heavier, and then the metals, of course, the metals affect all the catalyst units, but it will affect the earlier ones first, especially FCC. That is a deciding factor that catalyst poisoning. Garyville is a little bit more susceptible than Galveston Bay. Galveston Bay just has a lot of heavier processing units.

There's another company that starts with the V and that site, when it was BP that we're initially looking at much heavier crudes from Mexico that were going to come in, and that's what we're counting on. They would position themselves into that heavy crude. Just it's the metals in the Venezuela that is problematic and that will bring a constraint on to how much catalyst poisoning they have, and that depends on how much supply they're going to get because they have to use it a little bit quicker, that gets used up a little bit quicker. Then the coker itself, they got to make sure that they can get their coke drum cycle times down. It's an opportunity for them to reevaluate and then get the cycle times a little faster, but there's only so much speed you can get out of it.

CL [00:12:31]: Do you think the key cost driver that determines the economic viability of this is going to be the catalyst?

MB [00:12:46]: That would be that, yes, that constraint, but let me say that what is a mitigation to that is how you blend. If you can get some really heavy Venezuelan stuff and you've got a great supply of catalyst and you've worked it out and your hoppers will be full and you can sit there and have a three-month time, you're just running and cracking, and then you would blend that a little bit, and if your production guys planning can grab in a little bit lighter material to lighten up just a little bit, then you can get heavier stuff. You can get a multi-blend coming in if it's economically viable, by transport time, then you can pull in more Venezuelan heavy depends on that crack spread.

I should say the benefit of that is perfect source of that is west Texas Intermediate. Adding on to that, to what their benefit, what they can do, if they get a little bit too much lighter material, they have the ability to ship that inland to Catlettsburg, [which] is a good-sized refinery and Robinson is not too bad, but Catlettsburg is another one where they can get stuff out and get it over there so they can give themselves more volume across trying to bring in heavier stuff, not mess around the light stuff too much.

[00:14:26] What does bringing in more Venezuelan crude do to distillate yields, given the different properties?

MB [00:14:39]: The distillate yield isn't that bad. Let's just assume if you're making mid-barrel stuff, jet diesel, that kind of stuff, then your hydrocracker is going to be going, and that's fine. You can balance that with the coke drums, and this is where Galveston Bay just excels on having this flexibility to shift things around quite a bit. It's spaghetti works of piping and they can manage that manifold-type activity. Garyville is more linear. It's big, but it is more linear. It's got very strong distillation trains, but their flexibility is a little bit less than Galveston Bay. They can shift things back and forth between each other quite easily. When you say the distillate, usually, and they get these situations like this, they will go ahead and just run past nameplate and get a new, it's called MDR max distillation rate, and they'll just get a new in a 24-hour period, you supersede the last one, you get a new max rate, and that's what you set up for planning. Then you take those opportunities to push that hard. A great benefit of that is, Marathon is absolutely saturated with advanced process control. They're pushing the constraints to beyond prediction, and then Galveston Bay also has called GDOT, which is a dynamic optimisation technology that runs economic predictions also.

CL [00:16:32]: Does that shift towards more Venezuelan crude impact total throughput, which might be impacted by blending as well?

MB [00:16:49]: The Venezuelan crude could impact total throughput only if they're not prepared with enough catalyst because of the metals, then you have to pull back, and that's really pulling back on gasoline material and some other catalyst cracking units, and it's catalytic units, which is mostly gasoline range. Other than that, it's not, and then to try to keep that up, they could shift that to some other refineries that have some open source. I think they've got an AI tool try to balance those optimisations. Yes, theoretically, it can, but it can be mitigated also. Then the strain comes time. What's the transfer time to get these material out so you can unload and achieve more capacity?

[00:17:50] The industry has seen 9 million barrels per day go offline. How should we think about the relationship between the resulting high utilisation and practical throughput?

MB [00:18:26]: Absolutely. Yes, utilisation is the key, and that's every refinery. It's always been a major goal in Marathon, especially these two refineries, and the way you do that is with a good mechanical integrity programme. That has to be predictable. If you put more into mechanical integrity, then your utilisation has a better opportunity to be as high as it can. The things after that, and that's the biggest bang for the buck for utilisation. The other things that affect utilisation is, of course, weather or some acts of God that you can't do anything about, which are, of course, they're just there, but the other things could be, you could have some operational problem with personnel. There could be a union issue. All of these other issues around there. There could be some issues with the pipelines that transfer back and forth between the sites. Those little things come into it that are unknown hitters, but the mechanical integrity is by far the most important thing to achieve utilisation. You look at them and ask them, what is your mechanical integrity programme? How are you doing it? How are you predicting that? How do you move forward with it?

[00:19:47] Is there a rough degradation in utilisation that you would assign to a given month beyond the planned turnaround? If a turnaround is planned for October, is there a 1% degradation in utilisation monthly going forwards, or is that too tied to the maintenance quality?

MB [00:20:13]: That's true, and it is. That is baked into the turnaround because you're trying to, every refinery is trying to achieve at least seven years of turnaround now. That's the goal. If you can get 7.5 or eight, wonderful. You're planning your mechanical integrity, your cost spend analysis, all of that is planned towards at least seven years, but with that error, you could still have a 20% error within that and which means, to answer your question directly, it's not really a matter of days, it is months. The material doesn't degrade non-linear and quicker in those months. What will hit it is just the opportunities that you'll be missing. If you end up getting a turnaround in June, your year is in bad shape. That's why them changing for fall, that's great before winter, and then if it goes in January, it's not the end of the world at all. It's just a little bit off. Then there's a financial aspect. If it goes into spring, you better get things nailed down pretty darn quick. If it goes past spring, then you're losing a lot. It's more financial opportunity, lost profit opportunity, LPO, than it is seals are going to be breaking and all that kind of stuff.

CL [00:21:53]: If the turnaround got pushed into spring, how should we think about that from a quantity standpoint? Could you quantify that roughly in a range?

MB [00:22:08]: Yes. If you go into spring, the big hit on refinery is this unknown. You always account for 20-25% unknown things coming up. As you open things up and you have a cat cracking unit and you go in and you say, "Okay, we know we're going to have to drill and clean these things out," so you open it up and you expect it, and then you may find cracks in an air ring. You may find all kinds of little stuff. Your nozzles may be just broken or just flapping around. You plan that you're going to be able to repair certain things, but if you've got something where it's broken and your air ring laying on the bottom of the thing and nobody really knew that oddly, but it happens, then that could hit you and then you've got to go and buy material and get it going. If there's a delay like there was a few years ago in steel, you could be down many more months than expected. Then your decisions, how much of a Band-Aid do you put on it vs wait and fix. Spring just really ties your hands from that flexibility because you're missing driving season.

[00:23:32] How should we think about the gap between nameplate capacity and sustainable maximum throughput at Garyville and Galveston Bay? How much headroom do you think remains today, given that you pointed out earlier that Marathon is always trying to push it more?

MB [00:23:54]: Of course, nameplate is theoretical. Every refinery is built based on expected crude slates, so no refinery is exactly the same as any other, but that's the design rate. Then, as they've evolved, these sites are mature and have been getting knee replacements and hip replacements, just like everybody, heart valve replacements. They've been Six Million Dollar Man quite a bit, but that's where you get that sustainable throughput. That's where we're talking about the dynamic side of it. Now, the constraints that hit into that, and this is also weather-related, especially on these two refineries, cooling water becomes a major problem. You get in the summertime and your cooling water, it's just a little warm. That hits it. You can have problems with your hydrogen plants or hydrogen supply and that can hit it. You've got to keep that. That will help hurt that MDR. Other than that, the sulphur plants and cooling water, typically, you just run that, and I might be rambling a little bit, but if they see problems of sulphur plant, planning group goes and buys a little bit more sweeter crude and start blending that in to try to relieve the sulphur plant and achieve a little bit more nameplate.

CL [00:25:38]: Could you provide any approximations for the nameplate capacity?

MB [00:25:57]: Yes, they're always doing it. Normally, and I'm going to use the term MDR, normally that sustainability is if you can get, I don't know, about 3% or 4%, that's fantastic because especially in the big refinery, if you get 3% out of those guys, you keep pushing it up. When they do that, they're also on the process engineers analyse what it did to the pressure of distillation column, what it did to the differential pressure, how far did it push outside of that, even looking at what position were the valves on that thing and where were the heat constraints? They can look at that, they can say, "Okay, we got 2.3% more. This is great, raw, but we think we could have gotten 3.5%." Then the decision is made, do you push a little bit harder next week, or do you wait a little bit more time? It depends on the crude slate you have.

[00:27:03] Over the next one year, assuming cracks stay wide, what operational vulnerabilities still exist in the Gulf Coast system that you think may be overlooked by the market due to the strong crack spreads?

MB [00:27:26]: With strong crack spreads like this, they're going to take everything they can. It's a big competition to get crude slates with all your competitors. I'm saying that because the predicted financial value can have a little bit more error to it than you expect, just how well you negotiate. Exxon is trying to get the same stuff, so is Chevron, everybody else. Not so much Chevron, they're more upstream than downstream, but Valero is right there, and look at Motiva, the largest refinery in the US. You're competing against the same group cast of characters. There's a little bit of financial that could happen as a risk. Operationally, we're hoping for the Venezuela deal to come through. That's great. When it happens is a big constraint. That will relieve the Middle Eastern oil coming through, but all the pipelines that Saudi is doing and all of that, come by, if you get a little bit more, you can get some Kamchatkan oil. Nigerian Oso is always available, really light. The risk on it is not that great compared to the news. The idea that more Venezuelan means lower transportation cost and a higher amount that you have to get in up to a limit of that con-carbon content.

CL [00:29:03]: Could you elaborate on what you're seeing with the Nigerian Oso and Saudi pipelines?

MB [00:29:14]: Of course, Saudi is a sweeter and Nigerian Oso is almost gasoline coming out of the ground. You get these light material you can bring in, and if you get more Venezuelan, then, okay, they're going to bring in some of these lighter material also and have a certain blend around, and they're going to keep that on site. It's a dynamic blending. It's actually a kind of APC, you do a multi-period predictive blending. If you have these things on site, you can blend as you hit these constraints and keep pushing things that you're on. You've got a three-month window to get everything in under contract. This is the opportunity when you get Venezuela, and it's also an opportunity, not just for Venezuela in that area of the world, but opportunity for all the sweet, west Texas Intermediate, west Texas sweet, Saudi sweet, then like say, Oso, which is west African oil, that's a little sweet. It's an opportunity for all of those also because they need blend stocks.

CL [00:30:26]: What does the blend typically look like? Is it 70/30 Venezuelan and west Africa crude? What does the mix of Venezuelan crude that Marathon can handle look like?

MB [00:30:50]: Realistically, right now, if you get in some lighter material, you can get up to 80% of your crude, in Galveston, 80% of the crude could be Venezuelan. You could really capitalise on it. When they were building this, they weren't thinking of Venezuelan crude primarily, they were thinking of the Canadian sour. The pit mines there, they come out as bitumen, then you have to get it to flow and then they bring that down, so it was expecting more heavy Canadian crude in. There is not as much as it expected with the pipeline that didn't get completed, but there's quite a bit. With the Venezuelan in and the idea that you used to have a little bit more Mexican crude that was really heavy carbon content and lighter metals, then they were positioned to be this flexible. To answer your question, you can get up to 80% at Galveston Bay.

CL [00:31:56]: In Garyville, it's substantially lower?

MB [00:32:00]: It's a little bit lower, but not substantially lower, no, just a little bit. Garyville has, since they say it's a linear refinery, it's very predictable. What they can do on a Garyville, they can get this heavy stuff in and say, okay, we can run it like you're smoking a cheap cigar for, say, three weeks to a month, and then we need to run some lighter stuff to help clean everything up a little bit. You can dynamically balance that, but Garyville has that predictability in it that is there.

[00:32:35] Regarding margin capture, how much of that advantage can come from maximising the economics of the individual refinery vs optimising crude flows, intermediate streams and product placement?

MB [00:32:58]: I'm interpreting that is your crude flexibility and how you can export all the products from that. I don't know if that's what you said, but what I'm interpreting that question is that how flexible inside the refinery can you base your products to meet things with these heavier crudes. Some of those things are, geez, you can get your catalytic cracker, which we talked about. Other than logistics, you're getting outbound clean products, your amine units are going to be saturated and full, but not to any degradation of the crude slate. Then those off products are… Some of that are so light that they can be towards feedstock, towards clean fuel things. They're going to need some butanes and some other things, for biodiesel, some things like that which still need solvents. The offshoots of that, and I think you're talking about what the flexibility of that is, especially Galveston Bay, and a little bit of Garyville, but Galveston Bay has a little bit stronger way to get those materials out because of the Houston Ship Channel.

[00:34:38] Marathon achieved 112% margin capture in Q2 2026, some of that was driven by the company's ability to flex between jet fuel and diesel. How much further can it push jet yields before Marathon cannibalises diesel production to an economically detrimental degree?

MB [00:35:08]: Yes, you push in jetfuel, you basically cannibalise that diesel. You do push the fractionation of that kerosene-diesel boundary, which you make it more kerosene than diesel. There is some, the spec-slip penalties, they call that, and what's odd, there's a diminishing return where the jet has dramatically outweighs that diesel crack. Then when you hit that, it's called a cliff trend, and that cliff trend, when you hit it, it really drops off hard and it's by the process itself. You can get to the point where you can just go up to that point and then you're trying to justify that volumetric sacrifice. You're doing that, what's called the mid-distillate range. This is why that is less of an issue now is because Marathon at these two refineries, especially Galveston Bay, have pushed and become completely saturated with advanced control, which is a model predictive control technology. It's Aspen's DMC, and that actually helps you where you can push right exactly on that low-sulphur diesel pool and get your jet premium going and you can be almost at that distillation point that cliff trend, and actually operate on it.

CL [00:36:47]: By how much can Marathon realistically increase jet yield? I believe that, at Garyville, the company had a 3% increase in jet YoY because of the Robinson project and has increased capacity from 8% in 2024 to 12% in 2026. Is there still further runway without economic degradation?

MB [00:37:23]: What they've done is they've invested in stronger distillation capacity is really what they've done. What you can do with the existing columns is you retray them, and so you get high-velocity trays, and that's the heart of it. Now there is a little bit more on the cooling and the heat exchanger that they've done, a little bit more passes on, and some of those were changing some metallurgy increasing and getting a little bit better tubes and that kind of stuff, but the vast majority of that was retrain things and high-capacity trays. Then there's a give up on that. That high-capacity trays means you need to keep the thing full and pushed. If you're not, then when it gets a lower, if you have to limp along, it doesn't limp along. Those things do not limp along at all. That's betting that, okay, we're going to be really pushing max rates all the time in the future.

CL [00:38:26]: With the Robinson project increasing capacity, does that take away from diesel?

MB [00:38:32]: No, it's increasing total capacity. It's because Catlettsburg is landlocked, but Catlettsburg has the Ohio River. Robinson is right there in the middle of corn country, so they want a little more capacity. They want to be able to… A little more flexibility to shift between gasoline and diesel and mostly for farm equipment. Similar to Mandan. Mandan is a small refinery, but they provide most of the diesel from North Dakota.

[00:39:08] Despite the recent economics of renewable diesel, Marathon's management says it is not going to be allocating capital towards any new renewable diesel capacity. How much opportunity do you think remains today to increase renewable diesel output and improve those yields from existing facilities through small projects such as de-bottlenecking or process optimisation?

MB [00:39:49]: There is a limit on that, and realistically, any refinery corporate management is a little bit hesitant to put too much into it. That's not for any political or any kind of reason like that. The feedstock is very volatile. Even the three, and I think we're up to four Marathon, but those diesel plants getting feedstock, of course, they've invested a lot into the soy feeds that go all the way over to what was Martinez's refinery, which is all bio. Then in the Midwest, you got tallow, but those are regional, and in tall oil, you need a processing plant, soy, you need soy farms. That's why Nebraska and Iowa are great places for those things, but these de-bottlenecking have to happen the hydrotreaters, and there is an exothermic problem with that. Those types of light feeds have an exothermic constraint and within the hydrotreaters themselves. There's these fixed-bed reactors.

Oh, and another thing, these organics require an enormous amount of wastewater, so there's limits on that. The limits are a little bit more physical, which is why management go, how much money are we going to really invest in this compared to what the market is going to absorb. In my opinion, it seems like they're gobbling it up, but what's the exothermic difference between petroleum-based diesel and renewable diesel, and there's a little bit of exothermic differences in that quality gasoline. In my view, it comes into the same areas when we started whining about ethanol and gasoline. There's a small difference in it. Right now, on the renewables, it's an academic viewpoint rather than we have firm numbers from agencies like AAA did with ethanol and that kind of stuff and gasoline.

[00:42:16] In Q3 2026, do you think Marathon could increase its renewable diesel yield relative to Q2 and Q1 without particularly heavy investment?

MB [00:42:52]: I would say yes to a very small amount because any process, and you can even take that academically to design itself. Every design of the process units or even how the design units working together, it ends up being able to produce more or have more capability than the design, and in refining, especially that way. Not so much in chemicals and base chemicals, but in refining, there's always that way, because the material we're doing, we're really separating things by chemical temperature information. Thermodynamics is what we're separating these lump parameters. It's just a big goop of hydrocarbons that have a common thermodynamic profile, and we separate things like that and you get a little variability. It's like a bunch of first graders coming out, you got three different first-grade classes hitting the playground. They all swarm like birds, but they all have a little bit different personality and noise level. You can try to get it and corral it 3-4% any one way, but there is a limit on… Then that's where the other things we talked about come in a crude slate. Now, for renewable diesels, that crude slate is pretty much fixed. If you're built for tallow, you're tallow, if you're built for soya bean oil, that's what you got. It just becomes more of a, how much of this feedstock can you acquire?

[00:44:40] When the commercial team is requesting a yield shift, max diesel switching to max jet or vice versa, using Garyville as an example, how long does that take in practice to reach steady-state operations under the new parameters?

MB [00:45:03]: We always think and expect it's going to take two shifts, which is 24 hours, but realistically, in practice, it's a two-day affair.

CL [00:45:15]: It's a lot quicker than I would have thought.

MB [00:45:18]: Yes, and what makes it more predictable like that is the use of the APC. You already have a model built for that and then you switch it. It's really just operating the refinery. It's a little S-curve of that profile that it follows. It's not unheard of, you do that twice a year going between premium unleaded and regular unleaded gasoline also. It's just doing that with jet diesel is the little more units involved. It's a bigger volume of a problem than premium gasoline.

[00:45:59] Secondary products were noted as a bit of a headwind to margin capture, given the pricing lag relative to clean products. Operationally, what can be done to minimise the yield of low-value secondary products without overall crude throughput seeing degradation?

MB [00:46:27]: It's interesting. That goes to economics. Really, heart of a refinery, especially in that area, is hydrogen and how much hydrogen you get where you get it from. If you get that heartbeat in, then that's where you… Hydrogen is what you use to manipulate your molecules to do what you want with them and put them in the right place. Economically, when it hits a refinery, it's what the natural gas prices. If you look at Henry Hub, especially in that Gulf Coast region and see what that's doing, that's going to be the economic hit on the refinery, all of that stuff, and then that's going to decide, okay, is it worth it to push this harder, or is it worth it to make a little bit more other things?

[00:47:31] The Gulf Coast utilisation is at 100%. Marathon's management noted it's the lowest level of unplanned downtime this decade. Do you think it could continue to run at 100% utilisation without shutdowns? Is this sustainable and just a matter of getting better at managing it, or is that not practical or realistic for upcoming quarters?

MB [00:48:12]: Any time, and all of us, when we hear 100% of something, we always call it BS because we're not four. We know there's no 100%. When we say that, it's actually, you ask is when you say 100%, what timeframe are you talking about 100%? Because it's not seven years turnaround, 100%. That's not humanly possible, entropy in the world we live in, so what timeframe are you talking about? A year at 100%? If you're talking about two years, you got to, okay, there's a small test going on here. If they say 18 months, okay, explain that because what you're giving up with that 100%, you're giving up some reliability in call it fatigue. Geez, you get compressor seals, piping welds, furnace creep or tubes themselves, coking up things, rotating equipment, which come up, seals. The question on that is, say, 100% over a period of time and what are you expecting to have to do major repairs after that?

CL [00:49:28]: I was specifically referring to Q2 2026.

MB [00:49:32]: Yes. If they're going to do it 100% in a three-month period, I wouldn't doubt that. I would just expect to see them say we're planning a turnaround the following fall, and that's going to be a 2.5-month turnaround.

CL [00:49:54]: I haven't seen commentary about pushing turnarounds back yet, but some industry analysis suggests that refiners are delaying discretionary maintenance to capture these margins.

MB [00:50:13]: Oh, yes, and that is a strategy. Delay it, but what they'll do when they delay it is they'll go back and they'll look at their turnaround and say, let's accentuate more of the repairs we need, the major items, and let's see what are pet projects we need to scrape out of here because it's more about time than expense.

CL [00:50:37]: What are the early indicators to show that's the case of taking that too far?

MB [00:50:45]: I'd say most of those mechanical things are… They know where there's the pipes are doing water hammer. They know where that is. They know where the welds have some fatigue. They can even have that from the last time a weld was repaired somewhere because you already have tables, inspectors already have tables of how long this weld is going to last with things around this. There is an expectation that's predicted already to say how long it is. They're just saying what's going to make us roll the dice are these great margins. We got to capture the margins, so let's do a risk profile on capturing great margins vs having an unplanned unit problem. One unit may be a little bit more of a problem in Garyville than Galveston Bay.

Galveston Bay, they get one year and say, "Hey, we got a little bit more water hammer expected, so we're going to have to change some of these elbows out." They say, okay, what we can do on that unit because we can still run nameplate and shift things a little bit over or you can even take a major problem because you've got system-wide capacity, you can transfer things around. That's really in the heart of this 100% utilisation is where they can shift these problems when these gremlins start screaming a little bit. Marathon is really set up. It's a little bit by strategy, well, there's a lot by strategy, a little bit by luck, but it is set up, their system-wide capacity is what they're basing that utilisation on and then it is a really strong profile. There limits that system-wide capacity because you get place like El Paso is wonderful, but that's only good with the LA refinery. It's not so good for Catlettsburg and everybody else.

[00:52:57] Marathon's Q3 2026 turnaround activity is supposed to be concentrated mostly on conversion units in the Gulf Coast and Mid-Continent. How does maintenance on these specific conversion units impact the overall system's ability to upgrade products and capture margin?

MB [00:53:13]: They will stagger that between Garyville and Galveston Bay. They'll stagger those just because of the volumes those two refineries give, and you're talking about that with the Mid-Con units. Then there's a limited value that Robinson can bring. Catlettsburg is a pretty good-sized refinery. It's one-third the size of Galveston Bay, but its (inaudible 53.47) profile is pretty strong. Especially if they need to shift the lighter things to make gasoline range, then they can concentrate a little bit more on dropping some of that down into the distillate range. They're pretty good to be able to balance those two out. If everybody wanted to make a lot more gasoline like coming in the summer months, then that's where the constraint would come in because then everybody needs the amount of gasoline that's coming to heavy cat crackers in Galveston Bay and Garyville, they can't… Robinson and even Catlettsburg can't pick up an enormous amount of that value.

CL [00:54:32]: Where does the economic impact extend beyond the lost throughput of that individual unit while that turnaround activity is in place? Is there a bigger impact on the larger, heavier units on other parts of the refinery?

MB [00:55:38]: Well, yes, it is, especially since the refinery, it's almost like a bifurcation. It comes out of the crude unit itself and then everything is fed off of that. If you have… The crude unit is down, everybody's hosed, but if you've got… The coking units are going down and you can balance that and make more gasoline range material with the cat cracker, because then you have all the other reformers and all that are in place, and it just becomes a more traditional earlier refinery when we're using more gasoline, so it depends on which one. If you've got the cat cracker going down, then you've got to shift your production into that mid-barrel distillate, which is the diesel and the jet and all that, and you can't really deal with the lighter-range material. Then that's where they would shift that stuff off into Catlettsburg and Robinson and those kinds of guys. Then you'd still be able to keep your volumes up across the refinery. If that you mean with the lighter stuff like that, it becomes the big constraint when you have those lighter units just becomes your hydrogen supply, and that's where your problems start happening when you use up all your hydrogen supply, then you are limited.

[00:57:05] Refiners have to decide how much crude from Venezuela and WCS they want to flow through, and then Marathon has to blend that with some lighter crudes. Is there anything around that process of procuring, throughput impact or impact on yields that we should understand over the next 6-12 months as some of that potentially comes online? Where is the company's crack spread heading with respect to how much of a given crude slate it can take in? What does Marathon have to mix that with?

MB [00:58:17]: Especially Galveston Bay, such a wide spread that you can handle and I think we're talking about the slate itself, they get Venezuelan. What really pops up to them, let's say, they just max out and say we're going to get 79.8% in Venezuelan and we've got some WTI coming in, and we've got quite a bit of that. There's a little bit of Arab Light that we have available, so we're really going to push this for six weeks, then at the max limit. Then their real problem is going to be in the metals itself. They're going to be spending more on catalysts, two different kinds of catalysts, and when that comes in with the metals, then that will be their economic constraint and the crack constraint. Let's say, it's another time thing. How long are they going to plan to be doing that? I think the question on them, if they say we're going to take advantage of Venezuelan crude is really to say how much of it… There's 630,000 barrels per day. How much of that in the volume is that, and the real question is, how long do you expect to hold that value? Because then that's where it's a non-linear effect on when things get a little bit poisoned. You can go and have one or two beers are fine, but if you drink like a frat boy on a spring break, you're going to pay for it.

[01:00:03] What do you think is most important to keep an eye on regarding Marathon over the next 6-12 months, assuming the crack spreads stay wide?

MB [01:00:24]: Marathon, 6-12 months for them is, I think one thing you mentioned, I'll reiterate that is the turnaround schedule. The shift that into the winter months. Wait, let me ask, is this something you would be looking at to watch, or is this something you would be questioning them on?

CL [01:00:46]: Yes, something to keep an eye on over the next six months.

MB [01:00:54]: Really them is, I would look at the Henry Hub natural gas price, because if that thing goes up, even USD 1 per million BTU, that's going to affect them quite a bit. That is a good indicator. If that price goes up, then they're going to have to pull back on some of their heavy slates. If they think the price is pretty stable, they'll ride it and they'll keep going because the crack is more important, and something I always look at is, just in the industry, of course, and I'm sure it's old hat, you guys look at the VIX, the VIX, the volatility is going… I think the VIX today was 17 something. If you get over 13, okay, it's volatile. How much of the crude have they bought already to meet their three-month plan, and how secure is that?

Because if they go one year out, then they go 6-9 months, six months, and then three months, they're trying to be settled in and on what that slate is going to that refinery, and that's a rolling three months. If the VIX is, I look at the VIX, kind of high, it's 17, it's 20, then their profitability on that may be shifting and they may be changing their mind quite a bit, because that crude price is moving around and the spreads are moving around. If it's a little bit lower, it's more stable, then you can expect VIX, it's down to below 14, 13, they're going to be pushing that as hard as they can, and especially with the crack spreads, the yields are high, then it's just pedal to the metal. The constraint is really the Henry Hub price, what's hydrogen going for?

CL [01:02:53]: Okay. I think that's about all we have time for today, so I'm going to leave it there. I just want to say thank you for joining me and sharing your insights on Marathon Petroleum as well as the broader refining market.

MB [01:03:06]: Enjoyed it. Good conversation. Appreciate it.

Transcription ends at 01:03:10 of the recorded material