Good day, and thank you for standing by. Welcome to the Eos Strategic Outlook Conference Call. At this time, all participants are on a listen-only mode. After the speaker's presentation, there will be a question and answer session. Please be advised that today's conference is being recorded. I'd now like to hand the conference over to your speaker today, Ms. Elizabeth Higley, Director of Investor Relations. Please go ahead.
Good morning, everyone, and thank you for joining us for Eos' 2024 Strategic Outlook Call. Before we begin, allow me to provide a disclaimer regarding forward-looking statements. This call, including the Q&A portion of the call, may include forward-looking statements, including, but not limited to, current expectations with respect to future results and outlook for our company, technology and system performance, cost reduction initiatives, and market demand and trends. As well as statements regarding the conversion of backlog, LOI, and pipeline to revenue, the completion timing for Project AMAZE, the tax credits available to our customers or to Eos pursuant to the Inflation Reduction Act, or regarding our ability to secure financing, including final approval of a loan from the Department of Energy, LPO, which are subject to certain risks, uncertainties, and assumptions.
Should any of these risks materialize or should our assumptions prove to be incorrect, our actual results may differ materially from our expectation or those implied by these forward-looking statements. The risks, uncertainties that forward-looking statements are subject to, are described in our SEC filing. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made. We undertake no obligation to update any forward-looking statements made during this call to reflect events or circumstances after today, or to reflect new information or the incurrence of unanticipated events, except as required by law. The conference call will be available for replay via webcast through Eos' Investor Relations website at investors.eose.com. With that, I'll now turn the call over to Eos CEO, Joe Mastrangelo.
Thanks, Liz, and welcome to everyone to our strategic outlook meeting, where we're going to walk through our path to profitability. I want to focus on the agenda on page two. I'm going to go through the strategic framework for Eos, kind of refresh what we've been talking about over the last three years that we've been a public company, kind of update some numbers and talk about the direction that we're headed in. Then we're going to talk about the differentiated technology that we believe that we bring to market. And that's not just going to focus on a battery, it's going to focus on the battery itself, the system that we build, and the software that we use to operate that battery.
And then talk about the commercial momentum that we see building out in the marketplace as more and more people realize the need for long-duration energy storage. And then we'll wrap up with some financial views of how that, how those things fit into a path to profitability. So if we move to page three, I just want to. Today, we're going to, you know, you're used to hearing from Nathan and myself. Today, we want to bring in some of the other leaders that are driving key initiatives inside the company. You know, we're really proud of the team that we have. And we'll have Francis, Pranesh, and Daniel talk about differentiated technology. Marshall and Andy are going to talk about the building commercial momentum, and then Nathan and Jude will wrap up on our path to profitability.
Look, this is a small segment of the team that's doing the work here, but what I would say is, when you look at this group of individuals, broad and deep domain expertise across multiple companies in the energy space or industrial space. You can also see that, you know, we've refreshed the team here in the last, you know, 13-14 months to really get us focused on this next stage of development. I'm proud to have them talk to you today as we give everyone an update on where we're headed as a company. Now, with that, let me move to page four and talk about kind of what I would call the game board of how we think about growing the company profitably. It starts off with a really large TAM.
The amazing thing is, when you think about where we were in 2021 on projections out to 2030 versus where we are today, those projections have increased 30%-70%. So we're seeing higher demand. What we're seeing inside of that are larger projects and projects that require longer duration discharge. So really, when you think about what we're going to see evolving here over the next few years, is projects are going to get larger and larger as energy storage becomes a critical element in our energy mix. And we're going to require those systems to be able to do longer duration discharges, and also flexibility to allow it to really stabilize the total energy system.
Now, underneath this growth and these factors, you know, when we think about how will Eos win in the long term, it's going to be with differentiated product performance, building a simple system and getting that system to operate simplistically out in the field, and putting digital capabilities around those two factors to allow customers to maximize that asset. But in order to grow, you have to be able to have a highly efficient manufacturing capacity and a product that you can drive costs down to be able to deliver returns to investors.
So we think when you take the growth plus the technology mosaic that we've put together of a differentiated product, a simple system with digital capabilities, along with the manufacturing strategy that we put in place and the core tenets of the product that Francis will talk about, about how simple it is with widely available commodities. We believe when you put all that together, the company is now embarking on this journey to profitable growth, which we'll walk through today. So if we go to page five, I just want to take a second here and really drill down on total addressable market. Big numbers. When you look at this, you know, we've always said that the numbers are big, the company needs to grow into being able to capture the growth that's presented. So we've always been focused on core-...
Geographies as our overall opportunity. You know, while we're focused on the U.S., we're also dipping our toe in the water in places like the E.U. and U.K., India, Australia, and New Zealand. That represents a pretty large cumulative capacity for energy storage. But our true focus as a company is on the U.S. market, and you can see that the U.S., all durations, so that's anywhere from zero hours up to 24+ hours, is a large market. But where our technology excels is in the intraday market or in discharge times from anywhere from three-12 hours. That's a large segment that we continue to see growing and evolving.
When you think about how we've positioned the company for this, you know, we've talked about, and we'll get into more detail around Project AMAZE, adding 8 GWh of manufacturing capacity to deliver on those growth targets. That's around 4% of that total U.S. market. And at the same time, as Marshall and Andy go through the commercial momentum, our current opportunity pipeline today represents only 12%-18% of the U.S. storage market, and we believe that that demand is gonna continue to grow over time. So we think we have a product that fits a core need in the industry, and we're building the capacity to be able to deliver that product at scale and at cost, that will deliver returns for both customers and shareholders.
It's an exciting development, but you have to be able to deliver on that to be able to grow the company. We've often said that Eos is a company that has a highly efficient capital strategy for manufacturing. If we move to page six, we really wanted to take a moment to validate that statement. We did some comparative analysis on a select peer group, and we looked and found that in the last three years, we have one of the lowest CapEx spends of these companies in this space, and that low CapEx spend has delivered the highest number of cells. And then when you think about those two things put together and what we're trying to do with Project AMAZE, we really feel like we have something that we'll be able to grow upon.
And when you think about the manufacturing strategy that we put in place, discrete manufacturing into semi-automated manufacturing, into scaled state-of-the-art manufacturing, that's gonna allow us to be more efficient as we bring Project AMAZE online. We're learning from things that have happened in other industries as they've rolled out capacity expansions to help us avoid the pauses and overspends and things that we're seeing in the market, and we feel really good about where we are and the work we have to do. But there still is a lot of work left to bring our first state-of-the-art manufacturing line into production and then scale from there. So if you move to page seven, just wanna give a quick update on where we are in Project AMAZE. Pictures on the left are from ACRO's facility in Wisconsin.
You know, as we've talked about, we're doing our factory acceptance testing at ACRO's facility. It allows us to get the full benefit of their technology resources. And we've seen other companies that have tried to look at short cycling that and say, "Well, do your factory acceptance tests in your own facility." They've had challenges as they, as they've tried to do that. So ACRO is making progress in how they're doing that. You know, we are, you know, working hand-in-hand with them and really happy on where we are. A lot of work still left to do. Now, what we're trying to do, if you look at the right-hand side of this page, is implement the line faster. So we've added some additional controls, engineering resources.
We've come up with some prototype workhold part presentation and dunnage or prior moving material parts to be able to accelerate the development of the line as we've gone through this. So we've tried to find ways that allow us to get high-quality manufacturing, but do that a little bit faster. The next thing we're laser-focused on is delivering the project under budget. So where we are right now and what we've seen is that we believe the first line will deliver higher capacity at a lower projected CapEx or same projected CapEx spend that we had forecasted, with lower, significantly lower line one shakedown costs versus our plan. Now, we feel good about that number and where we stand today as we think about this because of the learnings that we've had on the semi-automated line.
When you look at the semi-automated line, the semi-automated line today, we've produced 57 cubes so far in Q4. We've seen our scrap rates go down significantly from when we last talked at the end of 3Q. You know, we were talking about scrap rates in the 5% range. Today, we're between 3%-4%, with December looking a little bit better than that. So we feel good about the learning that we've done. We feel good about what we have as far as knowing how to build the product, knowing how to flow that through the line in the semi-automated line, and how that transitions into the state-of-the-art line.
So as I said earlier, we're gonna do our factory acceptance test at ACRO to lower our execution risk, and then at the same time, we have incentives in place with all the suppliers to accelerate the implementation of that line. We're still forecasting the line to come on in 2Q. We'll keep everyone updated as we move forward, but I'll tell you, it's a proud moment when you look and you start seeing what we're doing manually now being done in an automated line, and you get excited about what that's gonna look like in our facility in Turtle Creek as we start to scale manufacturing. So great progress by the team. A lot more work to do, a lot of things we need to continue to manage and mitigate, but good progress so far to date with ACRO and the team.
If we go to slide eight, I wanna spend one moment before turning it over to the team, just talking about the product evolution. I think it's important to note that our chemistry has not changed, but the way that we manufacture our battery has. So if you take the pictures at the top of this page, Gen 1 was bolting together cells into a battery. Gen 2 started off with gluing cells together into a battery, and Gen 2.3 was bonding or using infrared welding to build batteries. Those batteries were all the same form factor, but what we learned in that experience was the battery was too big, the materials too costly, and the cycle time to be able to manufacture was too long.
So the team started working 14 months ago on the Eos Z3 battery, which is basically assembling the same core chemistry and manufacturing design of a battery in a new way of bringing that through production. So the battery continues to be non-flammable. We've taken manufacturing cycle time down by 120x from where we started just six years ago. Energy density, or watts per liter, is basically the amount of energy that you're getting out of your electrolyte in the battery. You can see that we've had a 3x improvement over time, and Francis is gonna talk about additional work that we're gonna be able to do to continue to improve that in the same form factor of the battery we have today, utilizing the same manufacturing equipment to achieve those improvements. We often get asked a lot of questions about round-trip efficiency.
You can see round-trip efficiency improving as we've gone through and developed the generations of our battery. The important thing here, and I think what I'd really like everyone to focus on, is as we talk about round-trip efficiency, when Andy Meserve talks about IRR on projects, we need to not just look at round-trip efficiency, but hold that in place and think about what that looks like with auxiliary loads to run the systems associated with keeping your system safe and operating. The same time, you have to think about when we're installing projects out in the field, we have a battery that doesn't have any degradation if it sits at zero charge. So there's a significant amount of cost that the customer gets to avoid as they're installing and commissioning systems. The next thing that we've also focused on is the self-discharge.
So imagine this: charge a battery up to 100%, it sits without discharging. What are you losing per hour of that charge as it sits there? We've improved this by-- we've reduced two-thirds of the self-discharge, so we're down below 1% per hour, which is significant for us and could open up additional use cases for us. Because one of the weaknesses we had in earlier generations was that the battery would self-discharge, and that was really driven by the form factor of the battery. Combined with the energy density, the battery was doing more self-discharge. So as we've taken the battery size down, lowered the form factor, and improved the energy density utilization, the battery is self-discharging less than it has in the past. Lastly, 20-year degradation.
So you're gonna see some numbers here that are improving over time, but seem high, but versus our peer group, are better. Now, some of this 9%-12% that you see with the Eos Z3, you size some of that into your system so that we don't have to do additional repowering as the project goes through its life cycle. And Andy will again talk about what that means as we operate the batteries out in the field. And lastly, we talked about no calendar aging, which is basically as the battery settles into its operating, the degradation remains flat, and the battery can sit at zero charge and not damage the performance of the battery over time. So this gives you a good framework of what we're talking about.
A high-growing TAM, projects becoming larger with longer durations, a differentiated technology portfolio, along with a scalable at high capital efficiency manufacturing process, with ability to continue to take cost out of the product, with a technology that's been improving over time over the last six years, with more room to grow, which Francis and the team will talk about. So with that, I'll turn it over to Francis, Pranesh, and Daniel to walk you through technology differentiation. Thanks for listening.
Thanks, Joe, and good morning, everyone. My name is Francis Richey. I've been with Eos for about nine years, and I currently lead the research and development team. I joined Eos a long time ago with a singular focus on designing, scaling up, and commercially launching the best possible battery and products, specifically for grid-scale energy storage. Building on Joe's introduction of the product, as many of you already know, the Eos Z3 battery is fundamentally different from previous Eos products in terms of the manufacturing process, manufacturing speed, and raw materials utilized in the battery.
As you can see in the image on the left-hand side of the page, the Z3 battery is assembled by inserting electrodes into a fully sealed bucket, designed to effectively eliminate external leaking while also increasing the speed at which modules can be manufactured, yielding a product that promises to be easier to build at scale. Listed on the right-hand side of the page are the four core materials used in the Z3 modules that have gone through extensive testing. Each of the core materials was chosen due to their reliable supply chain, as well as their use in other industries, speeding time to market and improving availability. The Z3 battery is the first Eos product to replace titanium current collectors with lower-cost conductive plastic current collectors, which have been in development with the technology team since 2019 and undergone extensive testing.
Conductive plastic eliminates the need for adhesives in the electrode assembly process and simplifies the electrode manufacturing process compared to titanium current collectors. In addition to conductive plastic, zinc bromide salts, graphite felt, and commodity plastics are the core materials in the module, and each can be recycled at end of life with minimal processing. While each of these widely available raw materials have been introduced into the battery module at various points in the evolution of the product. The underlying electrochemistry has not significantly changed since 2013, taking advantage of over 10 years of development and cycle life data.
Now, moving to page 11, I want to give a little bit more color on the history of the zinc-halide battery, and specifically, all of the changes Eos has made to the product since the invention of the technology, in order to better position the product for commercial success in the grid-scale energy storage market. On the left-hand column, you'll see a list of a few of the key characteristics of the earliest zinc-halide batteries developed in the early 1980s in a collaboration effort between Exxon and the Department of Energy. The challenges of designing, manufacturing, and operating cells at high pressure in a flow battery made the operating expenses increase in the levelized cost of storage, and also led to higher auxiliary energy consumption, meaning the flow battery was unlikely to be competitive with incumbent technologies.
All of these challenges led Eos to decide early in the product's evolution, to convert from a flow battery to a static battery design, eliminating the need for pumps and simplifying the operating expenses and the experience for the customer. After shifting to a static battery, it also became apparent to Eos that separators and membranes between the cathode and anode were not necessary in a static design, and the separator did not provide any inherent benefit to the electrochemistry, but added manufacturing complexity. Eliminating the separator significantly reduced cost of the product, but required Eos to engineer cell spacing, electrolyte properties, and electrode design to optimize performance in the static configuration. Now, moving down the list of characteristics on the left-hand side of the page, cell voltage in aqueous batteries is another important characteristic to consider for grid-scale energy storage.
The original battery was designed as a bromide redox cathode, yielding a cell charge voltage close to 1.8 V. But it is desirable to increase cell voltage, and Eos has achieved this by taking advantage of a mixed chloride bromide electrolyte, pushing charge voltage above 2 V per cell, and effectively reducing the number of cells needed to reach 1,500-volt systems by 15%. A significant improvement for aqueous batteries, while also reducing system cost. In the next line, you can see that the cathode substrate choice is also an important consideration. Cathode must maintain its structure and surface chemistry over the course of 20 years of operation across a wide range of temperatures.
We have moved away from the original carbon powder cathode to a graphite felt cathode, as graphitization has been shown to improve cycle life and reduce degradation across all operating temperatures in the field. Lastly, the current collection material choice was one of the few characteristics of the original battery design that Eos did not change. In 1983, the battery utilized conductive plastic current collectors, even though extrusion and injection molding of the conductive plastic proved to be very difficult at that time. Initial Eos batteries utilized titanium current collectors, but have since transitioned to conductive plastic current collectors, as the plastic extrusion and injection molding manufacturing technology has significantly improved in the past 10-20 years, which we believe will make manufacturing at scale more feasible and cost-effective.
After implementing all of the changes shown on this page over the past 10 years, I believe the battery chemistry design and manufacturing process is now well-positioned to take advantage of the grid-scale energy storage market. On the next page, I will discuss the steps we are taking to further improve the product. Flipping to page 12. We just discussed the invention of the product. Now, I want to discuss the further evolution. The focus of this page is to highlight a few of the projects on our roadmap designed to improve cell, module, and system energy density as we continue toward a product cost reduction goal. Volumetric energy density is an important metric for grid-scale energy storage, especially when considering that the cell and module costs are only a fraction of the entire system cost.
Fundamentally, this means that the objective of the technology team at Eos is twofold: One, to increase the energy we are able to store with the existing materials in the cell. Two, to further decrease the amount of space in both the module and the enclosure that is being occupied by non-active components that do not store energy. Within these objectives, the constraints are that the external dimension of the module and also the cube or enclosure should remain constant, and the current manufacturing processes should not be impacted. If we focus on the left-hand side of the page, we are outlining the key objectives related to increasing the energy we can store within our existing cell materials. You will see the cell, cathode, and electrolyte projects we are working on simultaneously to improve cell energy density.
Theoretical energy density of the zinc-halide battery is still well below the theoretical limit, and we believe the biggest factors limiting the ability to improve energy density are surface area of both the anode and cathode electrode, as well as the utilization of the electrolyte. The cathode felt, in particular, presents a great opportunity for improvement as the existing commercially available felts are not optimized for the battery industry, even though the carbon fiber manufacturing technology is currently available to improve the properties of the felt, specifically for batteries. The surface area of the felt can be improved by reducing the carbon fiber diameter, and we are working closely with our suppliers to manufacture felts with smaller fiber diameters under the assumption that higher surface area will increase cathode storage capacity.
Moving down the left-hand side of the page and focusing on the anode substrate, an advantage of conductive plastic current collectors over titanium current collectors is the ability to impart controlled and highly engineered surface morphologies during the manufacturing process. Titanium is a difficult material to modify, whereas conductive plastic can be more easily manipulated to impart surface texture and increase the surface area for zinc plating, effectively increasing the ability to plate thicker layers of zinc and increase the energy density of the battery. The benefit of this approach is that it does not require any new manufacturing processes, nor does it require our suppliers to change how the conductive plastic resin is manufactured, as the texturing is a process we manage on our manufacturing line. As the anode and cathode surface area improve, the electrolyte should also evolve to take advantage of these new properties.
We anticipate simplifying the electrolyte composition as we move forward, by reducing our additive package and increasing our active component utilization. After implementation of each of the cell material changes on the left-hand side of the page, we believe the cell energy density should increase by 10%-15%. Now, shifting to the right side of the page. In addition to cell material development, we're also focusing on projects designed to increase the energy density of the module, as well as the enclosure. Similar to other battery technologies, one of the critical routes to reducing product cost is related to the optimization of the active energy storage components within a battery module.
This means that the majority of the volume within the module, and also the enclosure, should ideally be occupied by the active electrode and electrolyte components, which actually store energy, rather than inactive components like support casing, sealing materials, welding and joining materials, and headspace materials, all of which reduce energy density and increase cost of the system. Eos is still early in the journey to increase the volumetric utilization of active components within our modules, compared to other technologies which have made significant strides in cell-to-pack volumetric optimization. Importantly, improving the volumetric utilization in the module does not necessarily require invention, and instead relies more on diligent engineering and continuous improvement of design and manufacturing as we move forward. Additionally, as the grid scale energy storage market shifts to longer discharge duration requirements, I believe the Eos product is especially well-suited to take advantage.
The discharge energy output of our modules and systems increases, moving from four hours of duration to 12 hours of duration. Lastly, the module packing density within the enclosure and ultimately the site energy density is another key area to reducing system costs. Opportunities exist to optimize system wiring design, BMS control architecture, and structural design of the enclosure itself. Daniel will be addressing some of our efforts on this topic in the next section. Taken together, the improvements represented on the right-hand side of this page, specifically related to the module and enclosure, are anticipated to further increase energy density by 5%-10% as we move forward.
Considering the opportunities outlined on this page, we feel confident in our ability to improve energy density of the product as we move forward, to help drive down the cost curve within the existing product paradigm, and without significantly impacting the existing manufacturing processes. Now, let me pass it to Daniel, who will talk about our simplified cube design.
Thanks, Francis. I am Daniel Chang, and I have the honor to represent product leadership. I joined Eos because I have an opportunity to make an impact to a full vertical battery product stack: cell, module, and system. Moving to page 13, I'm excited to talk to you about the Eos Z3 Cube, a product created with the mission of capitalizing on the value propositions of our Z3 battery technology that Francis just discussed. From our initial cube design, we worked with our key customers and took feedback on things that they value as projects get larger in scale and risks for project delays and monetization increased. The values that they told us they seek are listed in the top right. Our customers value a product that ideally requires simple project designs before it can be installed.
We aim to deliver this by considering our approach for cube-to-cube connections that is shown in the graphic to the left, and in detail A, where each cube only requires five connections that are installed by connectors and plugs. The next two values are interrelated and important to monetization of projects. Our cubes' auxiliary systems are fully integrated and pre-checked in the factory. This supports faster commissioning during construction, which accelerates time from install of our product to first discharge of the project. Now, let's do a deeper dive into these and other features that we have worked on, and the potential benefits to our customers for site development and engineering. The cube-to-cube connections allows us to have a continuous in-line arrangement for eight cubes. You can see this in the bottom left of your screen.
The cubes are connected to one another and to a marshaling box, which is a single location for site interconnections. The benefit of this layout is increased site utilization. While often mentioned as site density in the industry, the actual benefits include less roads to construct, less site area disturbed, needing stormwater mitigations, and even less fencing around the site. The single site interconnection is also utilized for auxiliary power feeds to our cubes. Auxiliary power is consumed for ventilation fans and control systems. Our technology does not require large auxiliary power consumers for air conditioning or redundant power for fire detection and suppression. So our customers only need a simple and single power feed run to each marshaling box. Let's now take a look at the bottom corner of the cube image. We have options on what is needed to support the cube.
Our customers have flexibility to use 4 or 6 support points, depending on soil conditions. Most importantly, when developing their sites, our product does not generally require a full concrete slab foundation, only a concrete beam or posts. This is an up to 80% reduction in concrete forming and pouring effort. Great, we have helped our customers to optimize the cost to get their site ready for our cubes. What is the next thing where we can make more positive impact? We look at construction costs for our system. If you look at our cube image here, our framing system permits lifting from the bottom using a forklift. This means the cube is moved off the delivery truck onto the supports that we just discussed without cranes.
Cranes capable of lifting the loads of our cubes costs a lot more than forklifts to secure and to operate, as it requires setup time before it can be used, and the crane needs to be disassembled after completing the project. Electrical installation is another area of cost savings during construction that we save from the connection feature that we discussed earlier. An estimated reduction of 19% is achieved because all the connections are external to the cube, and the labor required minimizes the need for skilled electrician labor, as it can be supplemented by general craft labor. Let's now talk about capital costs for battery storage projects. We have discussed optimizing product cost of the battery, site development cost, and construction cost. Other capital cost items that our system design minimizes include power cabling, which is reduced by half.
Land cost is also reduced by a site density improvement of 10%. Lastly, by making our cubes more compatible with industry-standard inverters, we provide additional supply chain options for our customers when they procure inverters for cost and supply lead time benefits. Takeaways from what you have just heard. We believe that product cost, land cost, site development, and installation costs incurred by our customers are all reduced through our simplified product design. This optimizes total capital cost spend to deploy our solution, which increases our customers' investment returns. Now, I will turn it over to Pranesh, who will walk you through our digital capability roadmap.
Thanks, Daniel. My name is Pranesh Rao, and I lead the product, system, and software engineering teams. I have worked in renewable energy and storage industries for many years. Recognizing the growing importance of long-duration storage, I jumped at the opportunity to join Eos a little over a year ago to help develop a safe and robust grid-scale battery solution, using a battery that is inherently well-suited to long-duration systems. In the last section, Daniel spoke about the simplicity of our Z3 Cube and its benefits in large-scale installations. Now, I want to talk to you about our digital roadmap and the advanced control system that we are building to overcome some of the major challenges in the industry that long-duration energy systems face in the field and provide a better performing product.
Starting off with the major challenges on the left, the first industry challenge is accurately calculating state of charge. The state of charge of a battery is a measure of how much energy is left in a battery. This is the equivalent of knowing how much gas you have in the tank of your car. It tells you how much energy you can get out of your battery, or in the case of your car, how far you can go. Imagine that you fill gas in your tank, your gas gauge tells you that you have half a tank of gas, and then you run out of gas a few minutes later because you are only really at 5% or 10%. Or imagine thinking you're empty, filling up gas, and then the gas overflows because you're really two-thirds full.
This is what happens with incorrect state of charge calculation. Now, suppose you have a fleet of cars, and you want them all to get someplace at the same time, using the same amount of fuel. This is known as system balance. It's where we want a set of battery modules to get to the same condition, charge or discharge, at the same time. With the state of charge inaccuracies that I mentioned before, and also depending on the way the batteries are operated, the system will become unbalanced over time. So the challenge is to know the state of charge of all the modules accurately and keep them balanced, so that we can get optimum performance from the batteries. The third challenge is working with batteries that require complex auxiliary systems like HVAC and fire suppression systems.
This adds another layer of complexity to the state of charge and system balance challenges that I mentioned before. Here at Eos, we are developing an advanced control system designed to address these challenges. As you can see in the center of the page, we're developing a sophisticated algorithm to calculate the state of charge. Using methods that are proven very useful in areas such as spacecraft control, we believe that our algorithm will help us calculate state of charge with much greater accuracy than traditional methods.... To address the challenge of system imbalance, we are developing what is called a state of health for modules. This is intended to provide us a measure of how balanced our battery modules are and let us know which modules need charging and discharging.
Going back to the gas tank analogy, the state of health is intended to let us know which gas tanks need to be filled when, and by how much, to make sure the entire fleet reaches its destination at the same time. The outputs of the state of charge and state of health algorithms feed into a machine learning engine that is designed to continuously improve our algorithms. The information from these algorithms will be used by our control system to drive a custom, scalable hardware platform that we believe will provide finer control over which batteries need charging and by how much. There are several advantages to the customer from the control system we are developing. Moving to the right of the page now, the first benefit is simplified control.
Since our batteries can operate over a wide temperature range and do not exhibit thermal runaway, they are non-flammable, and do not need complex fire suppression or HVAC systems, our auxiliary systems are greatly simplified. This simplification in control, along with the State of Charge and State of Health algorithms, should lead to increased revenue for the customer, since we should be able to maximize the extraction of energy stored in the batteries. We expect that this will also lead to reduced downtime, since the battery should be in a near constant state of balance, which means that they should not need to be taken offline to be balanced, as happens in many systems. In summary, we believe the control systems that we are developing will be simpler to use and will deliver increased revenue and reduced downtime to the customer.
They are intended to learn and get better over time and should provide a better storage solution to the market. With that, I will hand over the presentation to Marshall to provide a commercial update.
Thanks, Pranesh, and great to be with you all today. I'm Marshall Chapin, our Chief Customer Officer. I joined Eos nearly a year ago to run our sales, commercial operations, projects, and services divisions, and I couldn't be more excited to be a part of this team. It's a privilege to lead the engagement of our customers at this amazing inflection point, accelerating the world's energy transition with a long-duration storage solution that's shipping today. I'm excited to walk you through the details, but the key takeaways from me today are our pipeline is up $1.3 billion from last quarter. We're seeing deeper engagement after our LPO announcement and the Eos Z3 product launch, which has resulted in more than half of our pipeline being refreshed this year at higher prices, with utility customers bringing us larger projects and longer durations.
There's a lot to talk about here, so let me take you through each of those points in some more detail. As you look down the left-hand side of slide 16, we've seen an increase in customer engagement after the announcement of the Department of Energy's Loan Program Office conditional commitment and our successful Z3 launch. That's not surprising. That approval was the result of a highly rigorous two and half-year due diligence process with the DOE, and coupled with our Z3 launch, they act as important validations of our technology and our business. That validation has given our customers the confidence to dive deeper with us on their projects and ask for proposals. Those proposals are getting bigger.
After record-breaking amounts of new solar and wind projects in the U.S., we're now seeing an urgent need to support those intermittent renewable resources with large, long-duration battery storage, so they can perform like baseload resources and compete with traditional fossil fuels. As a result, projects over 1 GWh now represent $8.7 billion, or 2/3 of our total pipeline. As you can see in the lower left, 42% of our pipeline today is made up of utility projects. We're of course, still seeing strong developer growth, but utilities are moving faster than in previous years. ERCOT, which represents the Texas market, and California are both excellent markets for Eos, as those two geographies alone present great opportunity for growth, and many of our installations today in our pipeline are represented there.
We've won and installed projects with the California Energy Commission, and we're shipping Z3 cubes to ERCOT as we speak, making us well-positioned to win in California and Texas. We believe the successful implementation of the first manufactured line of Project AMAZE will provide us the credibility and capacity to deliver on these large projects. Now, let's move over to the four boxes on the right. Let me start with active proposals. More than half of our $12.9 billion pipeline comes from proposals delivered in 2023. Why is that important? That's the uptick in volume we've seen, driven by the passing of the IRA and Eos's advantage of the additional 10% domestic content kicker, because we're proudly a U.S. manufacturer, using predominantly U.S. raw materials, all of which are fully recyclable using today's standard methods.
And again, a big portion of these refreshed proposals are for large utility projects. Moving to the upper right box, from when we went public and pivoted to full commercial operations in 2021 to where we are today, the average selling price is up 40%, which is in line with the market. We believe we've earned that increase in price, in part due to our experience from field deployments like our project with Pine Gate Renewables in South Carolina, the largest non-lithium battery project deployed in the U.S. Now, let's talk about the timing of our customers' projects, specifically their COD dates. Because projects are getting bigger and because of the well-known complexities and challenges of the project development market and U.S. grid interconnect queues, our pipeline has a majority of projects, 60%, with COD dates in 2026 and beyond.
The great news here is we still have $5 billion in 2024 and 2025 COD projects, which should allow us to grow the company in the near term. Further, many of the larger 2026 COD projects have time to make decisions and could place POs in 2024, so we can manufacture in 2025 and commission and reach COD in 2026. The shape of our pipeline based on size and COD dates lines up very well with our plans for Project AMAZE and our manufacturing ramp, as Joe mentioned earlier in the presentation. Finally, the most important thing around all this, and what we've always talked about at Eos, the market is moving towards longer discharge durations.
I believe Eos makes a battery system that excels in those longer durations and provides the flexibility for multiple customer use cases as the needs of the energy markets evolve. I mentioned earlier our growth in California and Texas. Both markets are experiencing this shift. After record-breaking amounts of solar in California and solar and wind in Texas, both regions now see load drop way down during the day when the sun is shining and the wind is blowing. Then demand shoots back up as those renewables drop off and people come home from work. The shape of the resultant load curve was dubbed the Duck Curve, because the shape looked like the outline of a duck, where the duck's belly represented the daytime hours and its head, the evening hours.
But the belly effect is so extreme today, California's duck curve has been renamed the canyon curve, and this extreme effect means utilities need more long-duration storage to match the huge influx of renewables. In Texas, the ERCOT system has shifted from summer peak to a winter peak, often with durations extending between eight-10 hours. So in the two regions with the largest growth and demand for storage, the market has shifted to longer durations, where we believe Eos has a distinct technical and performance advantage over other technologies. That's why I'm excited to report that 64% of our pipeline today is made up of projects with durations over six hours, where the market historically has been focused on two-four hour durations.
The critical point here is that battery storage, and specifically long-duration battery storage like the Eos system, which can operate at peak performance even in durations up to 12 hours, is increasingly needed in these large markets like California and Texas, where the energy transition is gaining the greatest speed. So again, the key takeaways from me today are: our pipeline is up, both in terms of size and quality. We're seeing much deeper engagement after the LPO announcement and Z3 launch, which resulted in more than half of our pipeline being refreshed this year at higher prices, with utility customers bringing us larger projects and longer durations. I couldn't be more excited about our future. Now, let me hand it over to Andy Meserve, to talk through how the Eos solution delivers higher financial returns than competing technologies.
Thank you, Marshall, and good morning, everyone. My name is Andy Meserve, and I'm Vice President of Business Development here at Eos. I started my renewable energy career over 20 years ago, and I'm thrilled to be here at Eos for all the reasons we're discussing today. On this slide, I'm going to detail the Eos system's total value proposition. What I mean by that is digging into the realities of installing and operating our system and the substantial differences and advantages we believe we present over incumbent technologies. Almost all large energy projects are financed with a combination of project equity, tax equity, and possibly debt. Our customers look at all the variables of these projects to estimate a project's internal rate of return, which is typically being calculated over a 20-year period. Naturally, the greater the IRR, the greater the value of that project.
So let's dive into the major components of a battery project IRR, as this is the industry standard for measuring an energy project's value. The major components of IRR calculations are capital expenses, operational expenses, and revenues. When taking all these components into account, we believe the Eos solution provides our customers higher returns than incumbent technologies. The slide highlights how and where we're differentiated. Capital expenses are the total cost of the project, which includes all the development activities, equipment, engineering, and construction costs. The Eos system's density, which is the energy dispatched per square foot, is less than incumbent technologies, meaning we need a little bit more land to dispatch the same amount of energy, which equates to some additional civil work to install the system. That is where the Eos system adds some incremental costs.
As Daniel talked about, we're actively working to reduce these costs through our new enclosure design. Next, part of our customer's buying decision incorporates calculating the operational expense of the system. Meaning, what is the true cost and performance of the system? Not just the system's installed cost before it is operating. One of the most commonly quoted battery system metrics is round-trip efficiency, commonly referred to as RTE. This is the amount of energy discharged divided by the amount of energy used to charge. Eos' round-trip efficiency appears to be less than some incumbent technologies when solely looking at the values advertised on a system's data sheet. But you should not consider RTE without also comparing the auxiliary power consumption of that system throughout the 20+ year project term.
Because the Eos battery doesn't require expensive heating and air conditioning, as other systems do to operate, our auxiliary power consumption is 10 times less than that of other systems. As a household analogy, our system consumes the energy of a refrigerator, whereas other technologies consume the electricity of the entire house. It is important to understand that round-trip efficiency is really the total difference in the amount of energy dispatched and the amount of all the energy used to operate the entire system, not just the energy used to charge the system. At Eos, this is how we discuss and calculate round-trip efficiency with our customers. Our research on how other technologies discuss round-trip efficiency has found that the auxiliary power consumption of the HVAC systems has not been included in the published RTE values....
So after accurately calculating systems' round-trip efficiency and including auxiliary power consumption, the Eos system costs less to operate, resulting in a 100-200 basis point gain over incumbents. This delta becomes even greater in hotter climates, where air conditioning use is significantly increased, and today's largest energy storage markets in the country are California and Texas, so our advantage should be amplified in these markets. Another reality of the initial capital cost to install a non-Eos system is the fact that these systems need HVAC to be operating to prevent system damage prior to and during project construction. As these construction sites generally do not have access to grid electricity until the systems are fully installed and inspected, it is not uncommon to see expensive diesel generators used for this temporary power need.
By temporary, I mean a few months to a year, as large projects take time to construct and receive final permission to be connected to the grid. As Joe noted earlier, an Eos system does not require this costly pre-operational electricity and management. Our systems can sit and wait until the project is completed. Going back to the house analogy, imagine if you're having a new house built for you and your family, and during the entire construction process, you're paying for all the utilities like you would as if you lived in that house. That is what occurs with incumbent technologies. As Marshall highlighted, the energy markets are evolving towards longer dispatch periods as the Duck Curve becomes more extreme.
Our ability to be flexible with dispatch durations throughout the project term should allow our customers to participate in these revenue streams today, and as they evolve towards longer duration. The same Eos system can operate in three-1 2 hour durations without sacrificing any operational efficiencies. As you can see on the slide, we highlight the returns of a four-hour system and an eight-hour system to emphasize that our value increases as we move towards higher and longer dispatch periods. Why is this? And why wouldn't the IRR deltas be the same for different durations? This is because our chemistry dispatches more energy over longer durations than shorter ones from the same size system. As these durations become extended, you need fewer batteries to meet these longer direct dispatch durations. Incumbent technologies do not operate in the same fashion.
Shifting from a four-hour system to an eight-hour system means significantly increasing the system's size and operational costs or running at lower power, which will reduce efficiencies and further increase auxiliary load costs. Simply put, as the battery storage market continues to shift towards longer durations, the Eos differentiation improves. Now, let's discuss the longevity of Eos systems. Battery degradation is where the system's output is reduced based on time and battery usage. Most people have experienced this with their cell phones holding a shorter and shorter charge over time. The Eos system's extremely low degradation rate results in the system retaining roughly 97% of its capacity after 20 years of daily cycling, whereas incumbent technologies may lose up to 50% of their capacity during the same period. So traditional systems need to be repowered to maintain capacity.
Repowering means buying additional new batteries to add to a system throughout the project's life to try and maintain the system's initial capacity. Eos's consistent energy output and lack of repowering expense should provide up to a 250-350 basis point advantage over the incumbent. Many of our customers see this lack of degradation alone as a game-changing reason to use Eos. But again, even as this is a unique industry advantage, the sum of all of our differentials is truly the game changer. While we've discussed the key capital and operating expenses, we need to touch on battery system operations and maintenance procedures and costs. Not to diminish the importance of regularly scheduled service in our system.
The Eos system requires some air filter changes and on-site component inspections, whereas battery systems with HVAC, fire suppression components, and degrading battery cells require more expensive testing to determine which degraded cells need replacing, and HVAC and fire system services performed by third-party specialists. The service costs of incumbent systems are likely to be significantly greater than Eos for all of those reasons. Looking at all these capital and operational cost factors, prior to any potential incentives, we believe the Eos system delivers IRR gains in the 3%-6% range over incumbent systems, while significantly reducing real-world issues around safety and supply chain. On top of all that incremental value, Eos systems offer customers the opportunity to utilize the federal investment tax credit with domestic manufacturing adder for projects in the U.S. .
The additional 10% adder, above and beyond the standard 30% ITC, equates to a 100-300 basis points IRR gain. In most cases, this 40% ITC is applied to the entire project, not just the cost of the Eos battery. For example, a $50 million project with an Eos battery would receive a $20 million tax credit, where a system using a non-US manufactured battery would receive a $15 million tax credit. To summarize these components of the IRR, we believe the Eos solution delivers strong returns today and even stronger returns as the market shifts to longer duration. Now, let me turn it over to Jude and Nathan to discuss our path to profitability.
Thanks, Andy. My name is Jude Lepri. I joined Eos 15 months ago after an 18-year career in heavy manufacturing, and I am currently the leader of our FP&A team. I'm excited about the prospects of using that experience to launch a new product and transform Eos into a profitable company in the shadow of the heavy industrial past here in Turtle Creek. You have heard about all the great work that is going on at Eos. Now, I'd like to spend a few minutes tying those efforts to our cost projections. Moving to slide 19, I'm going to discuss our Z3 product cost walk from launch to scale, and how Eos has embraced a continuous improvement culture, which is highlighted by our cost and density improvement program, for which responsibility spans the organization.
Eos has a roadmap to reduce material and labor costs while increasing energy density that is expected to result in an 80% reduction on a dollar per kilowatt basis from launch to scale. Starting with the Z3 launch bar on the left, materials accounted for nearly half our product cost. Since we launched on a semi-automated line and volumes were low, direct labor was approximately 20%, and manufacturing overhead was another 30% of product cost. Looking at the at scale bar on the right, we anticipate material costs to be less than half the current BOM, and labor and overhead combined to be less than 1% of the launch cost. Since the Z3 launch in mid-2023, we have achieved over 30% of the planned cost reductions, with cut-in scheduled for early February.
Working the chart from left to right, R&D focuses efforts on increasing cell energy through increased modular and closure efficiency, as well as chemistry improvements related to our proprietary electrolyte solution. As Francis previously detailed, we have a multi-pronged strategy to increasing cell energy by reducing cathode felt fiber size, optimizing conductive plastic surface texturing, and eliminating electrolyte additives. Research is also identifying ways to obtain higher module and enclosure energy by reducing dimensions of non-active module components, increasing module packing density in the enclosure, and by increasing duration of discharge. Engineering continues to enhance product design to decrease product cost, as well as improve manufacturability, to optimize digital capabilities, and satisfy customer requirements while differentiating Eos Z3 in the market. As Pranesh and Daniel explained, our Eos Z3 Cube seeks to extend the value proposition of Eos Z3 battery technology through lower overall system costs.
Our customers benefit from simplified project design and installation, rapid construction commissioning, and accelerated time from install to first discharge. Eos designed and manufactured advanced control systems, reduced product costs, while enabling larger installations with simpler operation and system integration. Customer benefits may include increased revenue, reduced downtime, and simplified system controls. Moving to the next bar, supply chain is working to reduce costs through partnering with strategic suppliers, price negotiations on commodity materials, and insourcing of processes. The supply chain team is negotiating strategic agreements with resin, felt, and chemical manufacturers. For commodity products, we are securing the supply chain by identifying multiple sources and competitive bidding for costs. As we scale, we expect to experience deflation in our pricing. The supply chain team is also comparing costs of outsourced manufacturing against the cost of incremental investments and operational expenses to determine whether to in-source certain operations.
Finally, operations is automating manufacturing to reduce labor costs, finding ways to eliminate waste, and improving yields to reduce overall product costs. As Joe highlighted, ES expects to place the first state-of-the-art automated manufacturing line under Project AMAZE into service in Q2 2024, resulting in significantly lower labor costs and improved overhead cost absorption resulting from increased throughput. Once this plan is executed, we expect to deliver positive contribution margins, at which point we would accelerate manufacturing volume, investment, and growth. The previous slide outlined our product cost, which influences our contribution margin, overall cost of goods sold, and gross margin.
Focusing now on slide 20, we are going to discuss the costs necessary to support sales pipeline and backlog growth, to research increased density, to enhance product design for Z3 operating performance, to execute on the cost out initiatives that we just discussed, and to comply with the regulatory requirements related to being a publicly traded company. The slide displays 2023 projected and ongoing operating expenses, which include SG&A, as detailed on the left of the slide. Combined operating expenses are projected at just over $70 million. However, approximately $14 million of those expenses are non-cash in nature, which is primarily attributable to stock-based compensation and depreciation. At Eos, all employees are partially compensated with equity, as we believe this results in employees who are focused on the same long-term benefits as our shareholders.
In addition, there is another $4 million-$5 million of the overall expenses related to DOE due diligence, capital raises, and Z3 prototypes, which we believe will be non-recurring in nature. In total, over 25% of our operating expenses are either non-cash or expected to be non-recurring costs. Now, focusing on projected SG&A costs of $53 million on the left-hand chart, and working from top to bottom, we have the aforementioned non-cash stock-based compensation and depreciation expense. Next are the costs attributable to being a publicly held company, including increased costs for external audit and SEC counsel fees, Sarbanes-Oxley control implementation costs, and directors and officers insurance. Outside services includes things such as DOE due diligence fees, as well as capital raise and professional service expenses.
And finally, payroll costs, which includes the sales team, led by Marshall, responsible for selling and order growth, supply chain and operations, who are responsible for a large portion of the cost out, which I discussed before, and automation for Project AMAZE that Joe highlighted, finance and legal to ensure we meet our regulatory requirements, and HR, IT, and executive to ensure we have the people and systems to execute. Moving our focus to the right-hand side of the page, the annual projected R&D spend is approximately $18.5 million, which includes the R&D as well as product, system, software, and engineering teams. Again, breaking this down into a few key components, we begin with the non-cash stock-based compensation and depreciation, which I touched on earlier. Second, R&D outside service expense related to the cost necessary to patent and protect our intellectual property.
Third, materials used for Z3 prototypes, as well as testing of alternative materials to support cost out initiatives and increasing the energy density of our product. And finally, payroll costs, which are comprised from the R&D team performing the work to develop higher energy product that Francis discussed, and the engineering team to design the product for reduced cost and increased efficiency that Pranesh outlined. Across the bottom of the slide is a comparison of nine-month year-to-date operating expenses across the select peer group that Joe mentioned previously, who file on the same basis as Eos. The bottom row of data compares Eos versus the peer group net of stock-based compensation. Company A is over 2x Eos operating expense, and our closest peer, Company C's net operating expense, is nearly 24% higher.
As a team, we are focused on keeping our operating costs at the low level necessary to scale profitably while ensuring the infrastructure meets the requirements of being a public-operated company. We currently expect our operating expenses to remain relatively flat for 2024. With that, I'll now pass the call over to Nathan.
Jude, thanks for walking us through the product cost out and the OpEx build-up. Now I want to shift our focus to how we think about backlog and pipeline, and how it could convert to revenue over the next few years based on several potential pipeline conversion scenarios. You'll note that all of the scenarios presented include backlog as of 9/30/2023, combined with modest signed letter of intent or LOI assumptions from our pipeline. The scenarios for 2025 and 2026 also include a range of potential opportunity conversion scenarios and how those conversion scenarios impact the Project AMAZE implementation timeline.
Now, starting on the left-hand side of the page, I'd like to explain in a little more detail how these numbers are built up, beginning with what we have clear line of sight into based on our existing September 30th backlog of $539 million. In order to account for some of the long-term service revenue that will be realized over a longer period of time, as well as projects that could extend into 2027, we are only going to include 75% of our backlog in this analysis. Next, we map our existing backlog orders to the specific COD dates in the customer contracts, where available, or to the estimated delivery dates if COD dates are not yet finalized. Staying on the left-hand side, let's look at our LOI conversion assumptions.
We have assumed a 10% conversion rate of our current LOIs occurring over the next three years, which I'd like to point out is a little below our historical conversion experience. It's important to note that if LOI conversions come in below 10% in any particular year, we could accelerate backlog order production or vice versa, in order to optimize our capacity utilization. As we have discussed previously, LOIs generally represent the last stage in our pipeline before a deal gets into backlog. You should think about the LOI stage as a non-binding agreement that aligns our interests with the customers and has us on the same side of the table when pursuing and bidding on specific projects. If our customer wins, Eos wins. To clarify, LOIs never meet our criteria for a booked order and are not included in our backlog.
While the pipeline is not always linear, as some customers may skip the LOI and go directly to a binding contract, letters of intent can be an important factor in determining future customer contracts. Like any other industrial company, there can be uncertainty, and as we have explained in the past, we do not expect to convert 100% of LOIs to orders, but we certainly see them as an indicator of the future growth potential for the business. Focusing specifically on 2024, we see scenarios ranging between $75 million and $100 million based on our current production schedule, which includes us running either the semi-automated or the new state-of-the-art line at significantly less than nameplate capacity, while we're moving towards positive contribution margin.
As we've discussed previously, we are balancing critical customer deliveries and necessary training, product development, up until the point at which each individual battery module begins to contribute towards covering our fixed costs. Once we achieve positive contribution margin, currently anticipated in the fourth quarter, based on the combination of the cost out program that Jude walked us through, along with the customer opportunity conversion scenarios we just discussed, we intend to increase our production significantly as we continue to move towards positive free cash flow. Now, if you roll forward into 2025 and 2026, we apply the same methodology by looking at COD dates, both firm and estimated, in our existing backlog, and again, layer on a modest 10% LOI conversion assumption based on when customers have indicated they would like to begin receiving deliveries. However, that is not the entire story.
Let's shift our focus to the right-hand side of the page for a minute... Thinking back to what Joe outlined earlier, and if you believe where the market is headed and the size of the addressable market, we also need to factor in the potential pipeline conversion based upon the robust opportunity pipeline, which Marshall discussed earlier. So let's look at the remaining pipeline, excluding letters of intent, and apply three different conversion scenarios, ranging from 10% to 20%, which we believe is conservative, given the increased demand we've been seeing in the marketplace. On the low end, we're saying, "Look, if only 10% of our current pipeline converts, we see a path to $400 million and $1 billion in revenue in 2025 and 2026, respectively.
If 20% of that pipeline converts, we see the potential for $800 million growing to $2 billion in revenue in 2025 and 2026." So how does this align with the capacity expansion plans outlined in Project AMAZE? Looking at the check marks below each of the scenarios A, B, and C on the bottom right, you will see how this translates into the implementation of future manufacturing lines and how we think about our capacity expansion. We anticipate building the incremental lines when supported by customer orders and corresponding COD dates.
So if you do some simple math here, and you look at the endpoint in 2026, then apply a reasonable revenue rate to that, you can see what it takes to achieve these numbers with the 8 GWh of production capacity that we expect to have online when we have completed Project AMAZE. Again, these numbers are just based on the current opportunity within our pipeline today and do not factor in all of the work that our sales team is doing to pursue new business following the introduction of the new investment tax credits and the DOE's conditional commitment earlier this year. Any new pipeline growth that results from these efforts would be additive to the scenarios depicted here.
While not committing to specific revenue ranges for the next three years, the scenarios in front of us present opportunities that I believe position us for the future growth of the industry as funding, customer demand, and our capacity expansion align to take advantage of the larger projects and longer durations that we see increasing in our robust pipeline of opportunities. We intend to initiate 2024 guidance on our next earnings call. Now, moving to page 22, let's take a look at everything we've talked about today and recap where we are.
From the market opportunity and financial discipline that Joe outlined, to the ongoing product improvements and system enhancements that Francis, Daniel, and Pranesh walked us through, combined with our competitive advantages over alternative technologies and the exciting progress we are seeing on the build-out of the first state-of-the-art line, I believe that we are well-positioned to capture the longer-duration energy storage market opportunity that sits in front of us today. When you step back and look at a total addressable market that continues to increase and is currently expected to be over 500 GW hours by 2030, combined with Project AMAZE expectations of 8 GW hours of annual production capacity, our production would represent less than 2% of the total addressable market annually.
We've just talked about various potential scenarios and ways to look at how our backlog and pipeline could convert and contribute to revenue for 2024 and beyond. If you take some of the things that we've discussed today, along with some information from prior earnings calls, there is a 2024 scenario based on line one being implemented in Q2 and ramping to 35% production capacity in Q4, that could result in $75 million-$100 million in revenue for 2024. Moving over to cost out, the other component to gross margin is obviously costs. If you look at where Z3 costs were at the time we launched in the middle of this year, as Jude talked about, we see that 30% cost out has already been achieved and is expected to cut in in February.
In addition to that, we are actively pursuing an incremental 50% cost out, which we expect to be completed in the fourth quarter. The combination of revenue using average pipeline sales price assumptions, combined with the successful implementation of the detailed cost out program described by Jude, is expected to result in us achieving positive contribution margin in the fourth quarter of 2024. Positive contribution margin being defined as sales price, less direct labor and materials, taking into account the benefit of the production tax credits. Contingent on achieving this important milestone, we intend to increase production, and revenue numbers should begin to increase significantly because now every incremental unit that we produce is helping to pay down our fixed costs. Prior to that point, we will continue balancing capital requirements with critical customer deliveries and the ramp-up of the new line.
If you now look at the box in the bottom left-hand corner, as we have said for the last several quarters, we believe we have the cost base in place to support our future growth plans. As a result, we expect operating expenses, which include research, development, sales, and general and administrative expenses, to be flat year over year. I want to highlight, as you saw previously, we're actually the lowest in our select peer group when it comes to operating expenses. So if you take a step back, comparing ourselves to these other companies, we're delivering the highest revenue with the lowest CapEx spend and the lowest operating expenses. Now, there's still plenty of room for opportunity, but we plan to continue to focus on finding ways to ensure that we are spending capital efficiently as we operate and grow the business. Finally, let's look at Project AMAZE.
Where we're at here is that Project AMAZE, as we announced earlier this year, is a planned $500 million capacity expansion program, which is expected to ultimately yield 8 GWh of manufacturing capacity here in Turtle Creek. Sitting here today, with significant design and development work being done with ACRO on line one, along with the cycle times and the yield rates that we have been seeing on our semi-automated line, we expect line one to come in under budget, both from a CapEx perspective and, more importantly, on a shakedown or ramp-up and implementation costs. Given that each of our four anticipated lines are identical in nature, we think it's reasonable to anticipate that these cost savings would carry forward on subsequent lines, and as a result, we expect to come in significantly below the overall Project AMAZE budget.
In order to drive the right incentives, as Joe discussed, we have negotiated incentive payments with ACRO to support a potential acceleration of site acceptance testing and commissioning, as this implementation continues to be a key milestone on our path to profitability. Now, moving over to the right-hand side of the page, we do need to talk about capital. It takes capital to do everything we are talking about here today. Our focus from a capital raising standpoint is on getting the first state-of-the-art manufacturing line implemented. While you are aware that we have been utilizing the ATM to some degree, we continue to believe that there are multiple additional opportunities for capital available to us, and we have received several term sheets in recent weeks.
We look forward to sharing the results of our efforts in this regard at the appropriate time, but our current thinking is that this capital is likely to be in the form of structured debt or structured equity, or some combination of those two, in order to get us to the full implementation of the first state-of-the-art line. Looking ahead to future capital needs after the expected implementation of line one in the second quarter, we would anticipate the funding of the DOE loan to get us the capital we need to achieve profitability. Just to close out today, I want to thank everyone for their participation and look ahead to future events.
We expect to be back on the phone with you all in early March to discuss Q4 results, and then in the second quarter, we anticipate hosting an Investor Day to showcase our first state-of-the-art line here at our factory in Turtle Creek, Pennsylvania. With that, I'll turn the call over to the operator for questions. Operator, please open the line for questions.
Thank you. One moment, please. Standby for our first question. One moment. Our first question will come from Vincent Anderson of Stifel. Your line is open.
Yeah, thanks. Nice presentation. You know, good to see a thorough download on the heels of that new shelf filing. I think I just wanted to start with, with Joe to clarify on the ACRO progress. Great to hear you're below budget. You're shifting some spend towards speeding up implementation. I don't want to pin you down too much at this stage, but are those control engineering resources like a discovered necessity to stay on time or an incremental improvement over your original timeline?
It's incremental, Vincent, so we're spending the extra capital to try to hit the incentive plan that we have to accelerate the line in. You know, we've got to continue to work through this. You know, actually, Monday we're in ACRO all day to go through a review as we head into year-end, to then set up to see what the schedule looks like. Not ready to commit to a schedule, but continue to work it, just like we're working the cost side. You know, when you look at the cost side, Vincent, it's the CapEx coming in with a higher output on the line, the ramp cost coming in lower than we initially projected off of the learnings of the semi-automated line.
We're making great progress when you look at what's happening on the ACRO line, but we're, you know, not ready to commit to that accelerated schedule, but we're laser-focused on doing just that.
Great. And one for Francis, if he's available.
Sure.
But just... Great. Long time no see, Francis. Thinking about cell energy density, you know, usually there's a trade-off between surface area and cycle life. So I'm just, you know, curious, how much of a priority is that carbon felt? How much testing have you been able to conduct on smaller diameter fibers? And, you know, just thinking about the supply chain, is that something you might have to bring in-house? Is that outside, is that outside something you would, you would contemplate and just work with a strategic?
So, hey, Vincent, good, good question. Good to hear from you again. To answer the first part, you know, in other chemistries, there often is a trade-off between surface area and cycle life or degradation. Because our electrode materials are inactive, in other words, it's not an intercalation electrode, it's just a substrate that doesn't play a part in our chemistry. To answer your question on supply chain, you know, we've started procuring these materials. Historically, we've done a lot of work with different felt suppliers and different felt materials, and that's gonna continue through 2024. But we feel really confident in the-
... the trajectory we have to increase energy density based on the surface area of those felt materials.
And, and Vince, I think it's critical, like, when you look at Francis' page 12 and tie that into the cost outwalk that Jude laid out, you know, the critical thing on the cost outwalk, it's 80% down from launch. We've already achieved 30%, we've got 50% to go. Francis' page 12, which I think is important that everybody understand, is there's detailed programs in progress to take the cost out. We haven't baked in all six of those projects that Francis laid out, but all six of those are making positive progress. So any combination of those helps you get to your target.
The total opportunity set that we're looking at today, which I think over time the team will continue to generate, but the total opportunity set can get you up to 25% improvement in performance, and thereby take cost out and make the, make the offering more competitive. So it's a critical slide for people to understand that. When Nathan says contribution margin positive by the end of next year, there's a plan there. There's engineers working that plan. There's people out working with suppliers to deliver on that plan. It's real, and it's something that the team is focused on every day, and we've given the visibility into what we're doing to get to those numbers to achieve the ultimate goal of making Eos profitable.
That's, that's super helpful. If I could sneak one more in, and then I promise to pass it along. Just with regards to the balancing models that you're developing, you know, just curious, based on the data you've seen so far, are you comfortable with the hardware you've committed to on the BMS to support whatever model you ultimately optimize to? And just how does that play a role in the commercialization process if customers know or potential customers know you're still working on that?
That's a great question, Vincent. Yeah, we are very. The hardware we have selected is that we're working on, and the architecture we've selected is very scalable, it's very robust, and so we're very comfortable in being able to implement that and use it for balancing systems. So in terms of, you know, it because of the algorithms that we're using, the state of charge and the state of health, they're very customized to our batteries. The electronics as well ties in with that. So we are definitely very comfortable in how we're doing this.
We're also, you know, like as I said, our state of charge and state of health algorithms, you know, they tie into, machine learning algorithms to improve those, those algorithms, right? And so our electronics that we've picked will give us enough scope to expand that and scale in the future.
Yeah, and I think, Vincent, the other thing, just building off of Pranesh, right, that we should all realize is, assets out in the field operating make you smarter. You're never perfect, but you're smarter. So everything that's on that slide that Pranesh walked through, we're actually utilizing forms of that on the Gen 2.3. So the circles actually work between installed base and what's happening in the lab. You know, this isn't a—that middle part of his page is not necessarily a vision, it's what we're doing today. And what he's looking at now is how we scale that into the growth that we see in the company.
And I think it's very important to think about, like, as we go through cycle the battery and run, we then run it through his AI, and that AI then comes back and makes us smarter in how we run. The left-hand side of his page, when he was talking about state of charge and balancing, you know, everybody may think they understand how a battery works, but that's what's happening in the market today, and we've got to increase the utilization of the assets installed in the field. So what we're trying to do is get in front of that and being with the largest installed base, if you will, for non-lithium-ion technology, we're learning every day, making ourselves smarter, and it feeds into both Gen 2.3 and Z3. So Pranesh has this working.
What he's talking about is scaling into the growth that we see in the company.
All right, I appreciate it, everyone. I'll turn it over.
Thanks. All right, thank you.
Thank you. One moment, please, for our next question. One moment. Our next question will come from Chip Moore of Roth. Your line is open.
Hey, thanks. Thanks for taking the question. Wanted to ask one on more on the commercial side, customer conversations, I guess, and sort of comfort with your liquidity position, and then any update on those initial Z3 deployments, or are we still awaiting data on those?
Yeah, Chip, I'll start. Like, the Z3 that have shipped last quarter are waiting to be installed and commissioned to operate in the field. We are operating string systems here in both the factory in Edison, New Jersey, so we have operating data. On the question on liquidity, yeah, liquidity is a question, and I think the other thing, you know, as part of the strategy of the company and thinking about how all these pieces fit into a strategy, right? So the market is getting the demand is accelerating. Projects are getting bigger. Bigger projects mean that you have to have the capacity to be able to deliver. So the proof points of what we're doing in ACRO help people understand that.
At the same time, we walk customers through what we, what we've gone through today, and that we're trying to make it less expensive to get to the end goal by getting more efficient on the process of being able to do this. What customers are looking for is to say, "Look, if we're talking about 250 MWh project or 500 MWh project, we want to see the capacity in place." When you see those pictures that we have on the ACRO update, where they're actually moving parts and starting to do the work, it gets people comfortable. We've got to work through that. But when you think about where the overall industry is. There's permitting delays that we see across the board, and different delays in implementation, and larger projects take longer time when you start doing that.
I think one of the key things that Marshall talked about on his page is refresh and new proposals with larger, longer duration projects, and there's a cycle that you got to work through to get those projects through the process to be able to close. But we feel good about the conversations we're having, both with existing and new customers. I don't know, Marshall, if you want to add anything?
Yeah, no, just to note that, you know, customers are realizing that incumbent technologies actually have plenty of risk to go around as well, whether it be supply chain risk with rare earth raw materials coming from overseas or fire risk from thermal runaway. We're seeing customers come to us looking for a way to diversify their battery suppliers. So when we're given the opportunity to fully engage with customers and tell our story and show them the IRR benefits of our safe and tested technology, the message is resonating.
Yes, that's helpful, Joe, Marshall, thanks. And maybe 1, Nathan, I think you talked about line 1 coming in under budget and then some savings on shakedown. Is that - is $30 million still the sort of the estimate, or is this incremental savings on that? And then any way to help us think about how much is remaining in terms of spend there?
Yeah, so $30 million was the CapEx portion of the line. Nothing's changed from what we talked about on the third quarter on that front. When you look at if you take a step back and look at the Project AMAZE in totality, we announced a $500 million expansion program for the factory here in Turtle Creek. That included four fully automated lines, plus some additional projects related to supply chain and things like that. So roughly $100 million per line, that was both CapEx as well as the implementation costs, and then the ramp-up and those shakedown components. Joe walked us through, you know, we're seeing 40%, 50%, 60% improvements on the ramp-up and the shakedown, so that's a huge contributor to the under budget portion.
But, the CapEx piece is coming in slightly under budget as well. So we feel good about our ability to deliver on the first line below budget, and given that these lines are virtually identical, once you get the first one done, it's wash, rinse, and repeat, as customer demand requires it.
Yeah, and Chip, the only thing I would add on top of what Nathan said is, you know, when you're building product, actually building product, you learn. You know, and you look at the containers, the cubes that we've manufactured in 4Q, those down costs. It's not that we think we're gonna get there; it's that we know with what we're manufacturing that we can get there. We just have to be smart. And we've gotten a lot of questions about: Why do the FAT at ACRO? Why don't you just move it right into your factory and start there and save time? It doesn't save time. And we've seen other companies in the space try to do that, where they then had to scrap a line and start over again.
We're gonna be methodical, fast, but remember, we're actually learning because we're sitting in a conference room below the line-up above us, that's building batteries as we speak. As you learn, you revise, you update, and you keep moving forward.
You know, Chip, I should clarify one thing. When we talk about the total cost of a line, a lot of those costs are incurred after site acceptance testing and after the date that we anticipate receiving the first DOE advance. So I just wanted to clarify.
Yeah.
Don't jump to the conclusion that I need-
Yes
... to fund all of that upfront. That is, you know, again, when you think about this loan, it's based on eligible costs. Eligible costs are both the CapEx, the OpEx related to installing the line, but a significant portion is also those ramp-up and shakedown costs, and I just don't want us to get confused with the timeline there. A lot of that occurs after the anticipated first advance.
Great. Great clarification.
Yeah, that's good, good clarification. Maybe just a follow on there, just on, I guess, ramp-up. Joe, maybe for you, the biggest risks to get to that 35% capacity goal in Q4, I guess, you know, where do you stand on things like hiring and training, and how do you think about that?
Yeah, I mean, look, we've got, you know, multiple shifts on the semi-automated line. So when you think about the people we already have, and then moving them onto the automated line, one of the things, I think one of the things in Jude's page when he was talking about the cost out walk, you see that labor and overhead come down significantly, right? So if you think about two aspects of what we're doing right now, we've got two floors in a building. Upstairs is the semi-automated line, downstairs is where we're gonna put the automated line. So you think about our ability to use that overhead, we have doubled the overhead for the production that we have.
When you think about labor utilization coming down to that low number from where it is today, that labor utilization rate is just taking people, putting them on a line that has more throughput. So we have a lot of the people already working, doing things, and the good news is, if you learn how to do something manually, it's gonna make it easier when you do it as from state-of-the-art. So a lot of the people are in place, and we're working, and there'll be more to come on what we're doing here in the Mon Valley, working with the local community to recruit, train, retain, and grow the employees that we have and bring new people in. So, so we'll get there, and we've got time as you get into that 4Q ramp that Nathan's talked about.
What we don't wanna do, and I think we're trying to be prudent about how we do this, get the line in sooner, accelerate through that, get things up and running. But if you, if you just think that you can install a line and ramp up to full production, that's never happened in the history of manufacturing. So the plan ties to what we think is reality, with opportunity and risks associated with that. When you look at... people, Chip, it's a great question, but I think we're well positioned to be able to, to, to, to man that line as it comes, as it comes into operation.
Very good. All right, I'll hop back in queue. Thank you.
Thanks.
Thank you. One moment for our next question. Our next question will come from Thomas Boyes of TD Cowen. Your line is open.
Hey, thanks for taking the questions. I appreciate it. Maybe first, could you help contextualize the self-discharge improvements that you had on the slide? What is it for lithium-ion? I mean, it's great to see a marked improvement, but I just want to understand if there's also sort of a competitive advantage that you can lean in on.
Yeah. So, good question. You know, lithium-ion is still lower self-discharge than Eos. But we do use that self-discharge as an advantage for us because it does help us have a significantly lower degradation and longer lifetime compared to lithium-ion.
Got it. As a follow-up, maybe I would love to get your insight on, on kind of the potential for your technology to be paired with, with solar. With respect specifically to developers that are trying to capture the ITC credits, you know, for a project, given your significantly higher domestic content provisions, how, how's that dynamic playing out?
Hi, Andy Meserve here. So, yeah, no, our customers are absolutely pairing Eos with PV and, and some wind projects to, to capture the entire 40% of the domestic content kicker. So it's been very positive for us.
Remember, I think, Thomas, remember, you know, because of our domestic content level and because of that 40%, when you think of solar plus storage, utilizing Eos allows you to get... Depending on how you use your structural steel and other things, you can potentially get the whole project to be qualified just by using Eos today, and it's Eos today. It's not a forecasted build a factory, start from scratch, new product. It's Eos today qualifies and manufactures in the U.S., which allows people to get that full credit across their whole project, depending on how they source the rest of the scope of supply.
Right, and that's the to find a point on it, but that's without the solar actually having sufficient domestic content, correct?
The modules themselves, yes. Yeah.
Because we significantly exceed the 40% threshold for domestic content, you put that with some domestic steel, you should be able to push the entire project over the threshold, you know, if you plan properly.
Excellent. I'll hop back in queue. Thanks so much.
Thanks.
Thank you. Again, one moment, please, for our next question. Our next question will come from Christopher Souther of B. Riley Securities. Your line is open.
Hey, guys. Thanks for providing all the details here. Maybe we can just try to square the cost reduction slide with the kind of revenue slide, potential revenue scenario slide. You know, so if we're talking about, you know, next 18 months, you know, layering in those cost improvements with the revenue scenarios, you know, is that—you know, you mentioned contribution margin positive in the fourth quarter. Like, when does gross margin become positive, and what is kind of the, you know, mid-2025 or 2026, you know, kind of gross margin that you'd be kind of targeting, you know, at, at kind of full scale, I think might be helpful for folks.
Thanks. So just to recap a couple of things that we talked about today: cost out programs that are ongoing, you know, we see a line of sight, very clear line of sight into 80% cost out by the time we get to Z3 at scale. And given that we're talking here primarily about 2024, that's how we get to the Q4 numbers. When we look at the revenue side, again, just looking at the existing LOIs, the existing backlog, and the existing pipeline, and taking the revenue rates from that, those two lines converge. You know, similar to what we've talked about on previous quarters, those two lines converge to get us to positive contribution margin.
Once we get to positive contribution margin, as we talked about, we're only assuming that we're running the factory at 37% or 35% utilization for Q4, so there's opportunity to ramp that up further in 2025 and utilize a single line to pay down additional fixed price or fixed costs across the business. So once we get to that positive contribution margin, our focus then is on, okay, how do you get to overall positive gross margin and ultimately free cash flow? And we look forward to updating you guys with more specifics on that as we get closer.
Got it. Okay, so that 80% reduction is just, you know, basically through the end of next year, is the way to think about that?
Yeah.
Yes.
And that's-
Yeah.
That's to get the positive contribution. Okay.
Right. And Chris, remember, with a product that the life cycle—at its point in the life cycle where we are, that cost out will refresh itself as you go forward. You know, prior in my career, I've had hundred-year-old products that continue to drive costs out. So this is just up to the team that gave the ideas that we see today, you know, up with Francis, Pranesh, and Daniel, and the entire team working on this. There'll be more to come as we keep working and learning through this funnel.
Yeah.
Okay.
Ongoing. Ongoing, ongoing.
Excellent. No, that's really helpful. I'll hop in the queue. Thank you.
Thanks, Chris.
Thank you. I see no further questions in the queue. I would now like to turn the conference back to the CEO, Joseph Mastrangelo, for closing remarks.
Thanks, Chris. Thanks everyone for listening in. Look, a pretty densely packed outlook, and I think what I want to really do is focus on the strategy of the company, which we talked about in one of the earlier slides. A growing TAM, a TAM where the projects are becoming larger, durations coming longer, becoming longer, fitting into the overall technology of the company, that technology being differentiated from a product side. The things that Daniel talked about are extremely critical to make it easier for customers to use the product and take cost out in the field, combined with digital capability that will improve the assets as we look going forward.
I mean, that may be something that a lot of people think they know, but the important part is for everyone to understand that those are real projects identified, that we're working towards the chart that Jude talked about in the end. The capacity scaling, you saw the pictures. Those aren't promotional pictures. Those are pictures of Eos products going through their paces as ACRO starts to bring the line up. You know, so we feel good about being able to expand the capacity. We feel like we have the industry-leading capacity utilization or efficiency of that capacity when we bring it online. We feel like we're going to be able to do that more efficiently than what we forecasted back in the summer as the team continues to work. We feel like that all translates into a cost curve that brings us to be a profitable company.
You know, we've got the growth, and I think we have the plan, and we'll come back to update. But our goal as a leadership team is to continue to capture and position ourselves to win that growth, to deliver a solution that makes it easy for the customers to utilize the assets out in the field and deliver returns to shareholders. And we're laser focused on that, and we wanted everybody to go through some details to see that there's actual content underneath each one of those, to understand the work that's happening here in Eos every day. We'll continue to work on this and continue to update people as we have new things, but thanks everybody for listening in today. We really appreciate it.
This concludes today's conference call. Thank you all for participating. You may now disconnect and have a pleasant day.