Q2 2023 TPI Composites Inc Earnings Call
Participants
Jason Wegmann
Ryan D. Miller; CFO; TPI Composites, Inc.
Unidentified Company Representative
William E. Siwek; President, CEO & Director; TPI Composites, Inc.
Eric Andrew Stine; Senior Research Analyst; Craig-Hallum Capital Group LLC, Research Division
Joseph Amil Osha; MD & Senior Energy and Industrial Technology Analyst; Guggenheim Securities, LLC, Research Division
Julien Patrick Dumoulin-Smith; Director and Head of the US Power, Utilities & Alternative Energy Equity Research; BofA Securities, Research Division
Justin Lars Clare; MD & Senior Research Analyst; ROTH MKM Partners, LLC, Research Division
Mark Wesley Strouse; Alternative Energy and Applied & Emerging Technologies Analyst; JPMorgan Chase & Co, Research Division
Pavel S. Molchanov; MD & Energy Analyst; Raymond James & Associates, Inc., Research Division
Sherif Ehab Elmaghrabi; Research Analyst; BTIG, LLC, Research Division
Presentation
Operator
Hello and welcome to the TPI Composites 2Q 2023 Earnings Conference Call.
(Operator Instructions) Please note this event is being recorded.
I would now like to hand the conference over to your first speaker today, Mr. Jason Wegmann, investor relations. Please go ahead.
Jason Wegmann
Thank you, operator. I would like to welcome everyone to TPI Composites' Second Quarter 2023 Earnings Call.
We will be making forward-looking statements during this call that are subject to risks and uncertainties which could cause actual results to differ materially. A detailed discussion of applicable risks is included in our latest reports and filings with the Securities and Exchange Commission, which can be found on our website, tpicomposites.com. Today's presentation will include references to non-GAAP financial measures. You should refer to the information contained in the slides accompanying today's presentation for definitional information and reconciliations of historical non-GAAP measures to the comparable GAAP financial measures.
With that, let me turn the call over to Bill Siwek, TPI Composites' President and CEO.
William E. Siwek
Thanks, Jason. And good afternoon, everyone. Thank you for joining our call.
In addition to Jason, I'm here with Ryan Miller, our CFO. Jason is our new VP of Investor Relations and Sustainability; and is replacing Christian Edin, who has recently moved to Europe and is now part of our customer-facing commercial team. I want to thank Christian for progressing our investor relations program over the last 4-plus years and for his passion and leadership in advancing TPI's sustainability initiatives. Jason brings with him a wealth of experience in numerous financial leadership roles at a multinational aerospace and defense company. I want to welcome Jason to the team, and Ryan and I look forward to introducing him to all of you in the coming days and weeks.
With that, let's get to it. Please turn to Slide 5. To put it simply, it's been a tough quarter. Q2 has been challenging from both an industry and TPI perspective. And I can boil it down to 2 key issues: quality and volume, both of which I will discuss in a minute, but first, some good news. I am pleased to announce that TPI and GE have reached an agreement in principle to amend our existing supply agreement in Mexico to add 4 new lines to produce blades for GE's workhorse turbine in Juárez, with an initial term through 2025. TPI and GE expect to finalize this agreement in the third quarter.
Quality has been broadly discussed industry-wide over the past several quarters and quality issues have had a pronounced impact on performance, and we have not been immune to these issues. While the accelerated pace of new product introductions within the industry over the last 5 years and the push to get larger wind turbines to the market faster have significantly reduced the costs of wind energy, it is also a contributing factor to the wind turbine and blade manufacturing quality issues that have surfaced. As we reported last week, our financial results for the quarter were impacted by a warranty provision for the inspection and repair of blades primarily related to one blade type in one factory. We have responded to the evolving quality challenges with the following actions: We had a third-party complete an in-depth assessment of our existing quality system and are implementing improvement initiatives. We have and continue to engage with our customers more deeply and earlier in the design phase to minimize quality risks in the product design and manufacturing process. The good news is that our customers have significantly slowed the pace of new product introduction and recognize the benefits of standardization and industrialization.
We hired Neil Jones as our Chief Quality Officer, effective August 1, 2023. In this newly created position, Neil will oversee all quality processes, systems and controls relating to TPI's Wind business; and will report directly to me. Neil brings over 25 years of experience in quality and engineering positions in the wind and automotive industry. Neil spent more than 13 years with Vestas, in a variety of quality leadership roles, with the last 5 years as senior vice president, quality, health, safety and environment. Before joining Vestas, Neil spent over 20 years in the automotive industry, including engineering and quality leadership roles with TRW Automotive and a senior quality leadership role with Eaton Automotive. And finally, we are replacing certain members of our senior team to improve our operational leadership given the performance and quality challenges the company has experienced during the past year. We are confident that the steps we've taken will significantly reduce our warranty claim exposure going forward.
Now let's discuss volume. As we discussed in our First Quarter Earnings Call, we are still working on a handful of volume changes with our customers that will likely net out to lower sales for the year. Given the extended time it is taking to get clear guidance on the Inflation Reduction Act and the complexity of its implementation, ongoing challenges in the EU and the changing economics of certain markets, our customers continue (inaudible) on a market-by-market basis while considering existing inventory levels, all of which have resulted in volume changes from 3 of our 4 blade customers in 2023. In addition, there is no doubt that permitting, transmission, transmission queues, the ability of the broader wind industry supply chain to ramp volume, inflation and costs and the availability of capital are further factors impacting the timing of recovery, which we believe has likely pushed to 2025 as our customers continue to move transitions and new line start-ups to the right. Notwithstanding, we stand by our mid- to long-term sales and adjusted EBITDA targets we had introduced in our 2022 Year-end Earnings Call in February and are focused on positioning ourselves to deliver on those targets as volumes return and then accelerate to more robust levels the industry expects.
Now let's cover our overall Q2 results. Sales for the second quarter were $381 million and were negatively impacted by delivery delays of blades from increased inspection and repair activities. We also had lower automotive sales due primarily to lower bus body deliveries and field service sales were down as we had our field service technicians working on warranty-related efforts. Adjusted EBITDA was a loss of $38.9 million in the second quarter. As we discussed in our First Quarter Earnings Call, we expect the second quarter to be the low watermark for profitability for the year primarily due to annual wage increases kicking in, while incremental productivity benefits will be phased in over the second half of the year to offset those increases. However, at that time, we didn't expect the quality challenges I discussed a few minutes ago, which by far had the biggest impact on our adjusted EBITDA for the quarter. In addition to the warranty charge of $32.7 million we recorded, our adjusted EBITDA was also impacted by lower volume than expected, inflation and higher costs of inspection and repair activities.
Now I'll give you a quick update on the rest of our global operations, including service and automotive. Please turn to Slide 6. Notwithstanding the challenges we faced in Mexico during the quarter, our blade facilities in India and Turkey performed exceptionally well. Globally, we produced 661 sets and achieved a utilization rate of 85%. In global service, sales were down year-over-year due to a reduction in technicians deployed to revenue-generating projects. For the full year, we expect revenue to be down by about 30% year-over-year.
Things continue to progress nicely in our automotive business. However, we now anticipate automotive's 2023 full year revenue to be down from 2022 primarily due to lower bus body sales. We are also experiencing lower-than-expected sales in other automotive products due to our customer supply chain constraints and customer delays and transitions of new product launches. In the second half of this year, we are planning to launch 3 new automotive production programs. These programs include large structural panels for a commercial truck, a full battery enclosure also for a commercial truck and high-voltage battery pack thermal barriers for a light-duty truck.
Our customer diversification initiative is paying dividends, as 3 -- as these 3 launches are each with a different customer, with 2 of them being new to TPI. In addition, the products being launched show our investment in innovation and new manufacturing technologies are aligned with the needs of the automotive market. To support additional near-term growth, we have and are making additional capital investments in the light resin transfer molding, pultrusion and assembly processes. We expect our Rhode Island and Juárez automotive plants to be vertically integrated for these technologies by year-end, which will enable the scaling of our capacity and significant growth next year. We continue to explore strategic alternatives for the automotive business to enable us to scale faster and are encouraged by the initial discussions and expect to have more information to share by the end of Q3.
As it relates to our supply chain, the situation is significantly better than during the past 2 years. We continue to expect the overall costs of raw materials to trend down compared to 2022, while logistics costs have returned to pre-pandemic levels, both of which provide us some tailwinds for the back half of the year. Over the last couple of quarters, we've suggested that 2023 would be a transition year while the industry digests and/or waits for formal implementation guidance related to key components of the IRA in the U.S. and clarity around more robust policies in the EU. More and more, however, it's starting to look like the expected increase in volume related to the IRA and initiatives in the EU may not materialize broadly until closer to 2025.
Last month, the EU finally signed off on its renewable energy directive after months of negotiations. The emergency measures on permitting agreed last year will now become permanent. That means enforcing the principle that the expansion of wind is in the overriding public interest, applying a binding 2-year deadline to all permits and a population-based approach to biodiversity protection and requiring all EU countries to digitalize their permitting procedures. These new rules are an important step forward and will help unlock the 80 gigawatts of wind farms currently in the permitting pipeline across Europe. EU countries have until the middle of 2024 to implement them. Some are already doing so, for instance, Germany. And there it has led to an increase in permitting rates for [offshore] wind [and to] winning appeals against permits that were previously lost.
Here in the U.S., although guidance on many of the key provisions under the IRA have been issued, interpreting and getting clarification of the guidance will take some time. As you might expect with legislation as broad as the IRA, clarity around implementation will take some time. Just last week, the Federal Energy Regulatory Commission or FERC issued a long-awaited final rule on interconnecting generation and storage resources to the grid. Based on initial FERC statements, the final rule will implement among other reforms a "first ready, first served" cluster study, providing much needed relief for nearly 2 terawatts of renewables in storage that are currently waiting to interconnect. While we recognize the challenges the wind industry continues to face in the near term, with a continued focus on energy security and independence globally, we remain confident that demand for wind energy will strengthen once the current regulatory complexity is deciphered and global economies begin to stabilize.
With our current facility capacity of nearly 15 gigawatts, we expect our wind revenue to eclipse $2 billion, yielding a high single-digit adjusted EBITDA and a free cash flow percentage in the mid-single digits over the next couple of years. Today, we are operating 37 lines, including 4 lines for Nordex in Mexico. With transitions to larger blades, the start-up of new lines and the completion of the Nordex contract in Mexico in mid-2024, we plan to exit 2024 with 39 lines. These 39 lines will enable us to produce approximately 3,200 sets per year or 15 gigawatts. In IEA's updated net 0 by 2050 scenario, wind needs to reach over 400 gigawatts of installations per year with approximately 80% onshore and 20% offshore. Therefore, the market would have to be almost 5x larger than it was in 2022. 15 gigawatts of capacity will not be sufficient to meet the long-term needs of our customers, so strategically growing our global capacity and footprint over the next few years is a discussion we are engaged in today with all of our customers, as we are in a unique position to capitalize on the growth in the wind industry.
With that, let me turn the call over to Ryan to review our financial results.
Ryan D. Miller
Thanks, Bill.
Please turn to Slide 8. All comparisons discussed today will be on a year-over-year basis for continuing operations compared to the same period in 2022. Please note our prior year financial information has been restated to exclude the discontinued operations from our Asia reporting segment, as we shut down operations in China at the end of 2022.
In the second quarter of 2023, net sales were $381.3 million compared to $392.5 million for the same period in 2022, a decrease of 2.9%. Net sales of wind blades, tooling and other wind-related sales, which hereafter I'll refer to as just Wind sales, decreased by $4.9 million in the second quarter of 2023 or 1.3% compared to the same period in 2022. The decrease in net sales of Wind during the second quarter was primarily due to a 2% decrease in the number of wind blades produced due to lower customer demand and delivery delays [from increased] inspection and repair activities, a decrease in other wind-related sales for mold decommissioning services and lower average sales prices due to the impact of raw material and logistic cost reductions on our blade prices. These decreases were partially offset by favorable foreign currency fluctuations and an increase in tooling sales. Additionally, our utilization in the second quarter of 2023 was 85% compared to utilization of 88% in the second quarter of last year.
Field service sales decreased by $2.9 million in the second quarter compared to the same period in 2022. The decrease was due to a reduction in technicians deployed on revenue-generating projects due to an increase in time spent on nonrevenue-generating inspection and repair. Automotive sales decreased by $3.4 million in the second quarter compared to the same period in 2022. The decrease was primarily due to a reduction in the number of composite bus bodies produced and a decrease in sales of other automotive products due to our customers' supply chain constraints and customer delays in transitions of new product launches, partially offset by an increase in fees associated with minimum volume commitments.
Net loss attributable to common stockholders from continuing operations was $80.8 million in the second quarter 2023 compared to a net loss of $25.3 million in the same period in 2022.
Adjusted EBITDA for the second quarter of 2023 was a loss of $38.9 million compared to adjusted EBITDA of $5.6 million during the same period in 2022. The decrease in adjusted EBITDA during the second quarter ended June 30, 2023, was primarily due to increased warranty costs, higher production costs for additional quality control measures implemented at certain manufacturing facilities and increased labor costs in Turkey and Mexico, partially offset by foreign currency fluctuations, cost savings initiatives and lower start-up and transition costs.
Moving to Slide 9. We ended the quarter with $170 million of unrestricted cash and cash equivalents and $195 million of debt. We had positive free cash flow of $6.2 million in the second quarter of 2023 compared to $19.4 million in the same period in 2022. In light of a challenging quarter, we placed a significant focus on preserving cash and ensuring we efficiently deployed our working capital to make sure we can comfortably execute key initiatives as we move forward. We do expect a modest level of cash burn during the second half of the year as we satisfy our warranty commitments and continue to implement quality improvement initiatives. Note that most of this cash burn is expected to occur in the third quarter.
I know many of you want to understand how we are thinking about our cash position beyond 2023 and as we enter a period of time in 2024 when we expect to be starting up idle lines and transitioning to longer blades. The bottom line is we continue to be confident in our liquidity position. To shed a little color on the moving pieces: In 2024, we expect positive EBITDA and working capital initiatives to be our primary sources of cash. We are expecting EBITDA to significantly improve in 2024, compared to 2023, as we get the cost-of-quality issues behind us and we effectively [shed] the losses from our Nordex Matamoros plant. We also expect to get some advances from our customers to support the ramp of idle lines. Offsetting these sources of cash will be CapEx primarily related to the transitions and start-up of idle lines, interest and taxes as well as cash payments to Oaktree for the preferred dividends. Our first dividend payment to Oaktree will be in January 2024. In total, we expect to make about $40 million in payments for preferred dividends to Oaktree in 2024.
End of the day, we believe our balance sheet, our projected liquidity position and our operating results will enable us to navigate the short-term challenges and invest in our business to achieve our mid- to long-term growth targets of $2 billion-plus in Wind sales and high single-digit adjusted EBITDA margins.
Moving on to Slide 10. As a result of the quality and volume changes Bill discussed, we're updating our financial guidance for the year. Sales are now expected to be down by about $100 million at the midpoint of the ranges from our initial guidance. Approximately half of the reduction relates to lower customer demand for blades and delays from inspection and repair activity. About 1/4 of the reduction relates to lower field service sales, as technicians have been diverted to nonrevenue-generating work. The remainder of the reduction relates primarily to lower ASPs from supply chain reductions and lower automotive sales than expected.
We do continue to expect to achieve low single-digit adjusted EBITDA margins over the second half of the year, which is consistent with our initial adjusted EBITDA guidance for the full year. However, with the loss from the second quarter, we now expect our adjusted EBITDA for the full year to be a slight loss of less than 1% of sales. With the reduction in sales volumes, we now expect utilization to be in the low to mid-80s versus our initial guidance in the mid- to upper 80s.
With that, I'll turn the call back over to Bill.
William E. Siwek
Thanks, Ryan.
Please turn to Slide 12. We remain very bullish on the energy transition and believe we will continue to play a vital role in the pace and ultimate success of the transition. We remain focused on managing our business through the short-term challenges in the industry and are excited about how we are positioned to capitalize on the significant growth the industry expects in the coming years.
I want to thank all of our TPI associates once again for their commitment, dedication and loyalty to TPI.
I'll now turn it back to the operator to open the call for questions.
Question and Answer Session
Operator
(Operator Instructions) Your first question comes from Julien Dumoulin-Smith from Bank of America.
Julien Patrick Dumoulin-Smith
Sorry, guys. Can you hear me? I was on mute.
William E. Siwek
Yes, got you, Julien.
Julien Patrick Dumoulin-Smith
Just a couple of different things here. So look. I know the backdrop has been challenging here of late, but can we talk about the quality-related matters in a little bit more detail, right? So you disclosed here a issue with a customer at a given site under warranty, but you also discussed a third-party independent evaluator has come through, reviewed your operations, et cetera. Can you discuss what the third-party review found? And have you needed to pursue quality changes at other sites and/or for other customers thus far, if we can kind of get into some of the other permutations here, if you don't mind.
William E. Siwek
Sure, yes. We brought in -- actually we brought in the third-party quality review. Or that was our choice. Our customers didn't know we were doing it, quite frankly, at the time. I commissioned that a couple of months ago. And it was more just to make sure that, as -- again, we were seeing more issues within the industry. And we had a little bit of an increase in what we would call nonconformances in some of the blades we were producing, so wanted to get a very objective third-party view of our quality management systems, what we were doing well, what we needed to improve on. So I would tell you it's not -- I mean in virtually all of our plants we're in actually really good shape. We just had a couple of facilities that were not where we would like them to be today. So that's kind of the genesis of the findings. There's improvements we can always make, and we're making those, but overall there was nothing that was of major concern. It was just improvement on what we had already done. So that's, I think, the -- that's about the third-party reviewer, but again we have -- they're -- just so you guys know: I mean blades are incredibly [manned], right? I mean it's anywhere from 1,500 to 3,000 or 3,500 direct labor hours per blade, hundreds of layers of glass laid up by hand, infusion that is somewhat automated but still relatively manual, as well as laying core down, et cetera, et cetera, so there's a lot of manual processes that go into it. Every one of our customers' blades is different and each of our customers' blade models are different, so it's not like it's a highly automated process, which then leads to more room for nonconformances when you're building a blade that can be 80 meters long, 30,000-plus pounds, right? And so we -- the whole industry has work to do, especially as we move from relatively moderate-sized blades to very large blades very quickly. The challenge is getting up to serial production. Once you get to serial production, the quality is obviously much better, but when you're switching blade models as quick as you are, you never really get there, so the good news is, with NPIs slowing down, I think our -- any improvements we're making from our quality standpoint -- I think the instance of quality issues going forward will moderate back to levels that we used to see which were very low.
Julien Patrick Dumoulin-Smith
Got it. And just to clarify this: You said there was a couple sites with issues. Just can you elaborate a little bit on what's been done to remediate? And it seems like you didn't elect to pursue a warranty claim on that second site. If you can elaborate...
William E. Siwek
So Julien, we evaluate our potential warranty claims continuously. There are some other small warranty claims out there. I mean we've got a fairly sizable reserve now, so we feel very comfortable that what we've got is covered. So I mean the other -- what have we done to protect them? We have firewalls that we put in place. And again remember we have -- our customers are in our plants with us. And so not only are we inspecting them, but generally our customers are inspecting them as well before they get shipped. We've increased the level of inspection. We've increased the numbers of inspection during the production process, not just at the end of the process. And in some cases, we've invited third parties into our plants to validate the inspection work that we've done. So...
Julien Patrick Dumoulin-Smith
Right -- and sorry. Lastly, you said that you have work to do -- go for...
William E. Siwek
Yes, you always have work to do, right? It's called continuous improvement. And again when it's you're -- when you're dealing in an extremely manual process, there's always improvements you can make, but we've had a very, very good long-term track record from a quality standpoint. This large -- the large issue that we specifically provided for this quarter, think about how we talk about it. It's inspection and repair, so a lot of this is proactive. There were some nonconformances found out in the field before blades were put up tower. We looked at those with our customers and we developed a plan to make sure that we inspected every blade. And if there was the need to repair, we would repair, and if not, we would move on, so it's not like we have blades failing in the field, blades flying off towers. This was more proactive to make sure that we never get into that situation. And again, for many of these blades, there'll be relatively minor repairs that are done. Some blades will likely not have any and others will have some repairs to do before they go up tower.
Julien Patrick Dumoulin-Smith
Right. And you still reconfirmed your high single-digit EBITDA margin target, although you didn't specify a time line necessarily, but it seems like the warranty cost, the higher OpEx, et cetera still hasn't taken you off that goal.
William E. Siwek
Well, that higher single digit, we said in a couple years, right? We've got to get through -- we've talked about '23 being a transition year; '24 a bit of a transition year, as some of the transitions that we planned from a line standpoint as well as start-ups have moved to the right a bit by our customers. And so we see '23 and '24 really as not only industry transition years but TPI -- some TPI transitions as it relates to lines; and so we see ourselves exiting '24, entering '25 more on that run rate.
Operator
Your next question comes from Mark Strouse from JPMorgan.
Mark Wesley Strouse
I'm thinking back to the second half of last year, maybe around the same time that you announced the GE contract in Iowa. You were talking about in -- you've been in discussions with other OEMs, as far as potential U.S. manufacturing. Has this quality issue impacted those conversations at all?
William E. Siwek
No. We're still -- we're evaluating sites today and still in discussions with multiple customers for sites within the U.S.
Mark Wesley Strouse
Okay, good to hear. And then with the announcement today with GE, how should we think about when -- assuming that you get the full contract in 3Q, what the timing of that might look like, as far as when that might start? And since it's in Juárez, is there any disruption to your existing lines while that process is ongoing?
William E. Siwek
Yes, so if you remember, Mark, we had a -- or we had an empty facility in Juárez. So this is the empty facility. So GE will be moving into that new facility and we expect production to start early next year.
Operator
Yes. Your next question comes from Justin Clare from ROTH MKM.
Justin Lars Clare
So I wanted to start with the demand picture a bit here. You indicated demand might be pushing to the right a bit and into 2025. Just given that backdrop, maybe you can walk through what your expectations are for your open lines. It sounds like you're going to be exiting 2024 with 39 lines, so it seems like you're adding the 4 with GE. There's 2 others that you'll contract. And then maybe the 5 lines in -- or the potential 5 lines in Iowa, those might not get contracted in the near term. So just maybe if you can walk through what you're expecting here.
William E. Siwek
Yes, yes. So as -- the line count was a little bit different last quarter. And that's because, as we've gotten deeper into discussions with customers and we looked at blade sizes, the number of lines have actually come down a little bit, but if you were listening to the prepared remarks: We wind up with the same number of gigawatts of -- or the same capacity at 15 gigawatts because the larger lines will generate -- obviously we can produce the same amount of gigawatts -- or megawatts that we would have otherwise, so we add 4 lines in Mexico. We subtract 4 lines in Mexico as well, the Nordex lines, right? And then we're -- you think about Iowa, which is probably 4 to 5 lines at this point; and then India, adding a couple of lines there as well.
Justin Lars Clare
Okay, got you...
William E. Siwek
Do you want me to give you the exact -- we're -- yes, we're at 37. You add 4 in Iowa potentially. Again all -- GE -- we're working with GE to still determine the right time for that to open based on clarification around some of the IRA and what their demand needs are. So that's still to be determined, but think of it as 4 lines; Mexico too we just talked about, which is 4 lines; 2 more in India, 2 lines in India. We take out the 6 -- or the 4 lines for Nordex in Matamoros. When we transition to a larger blade in another plant in Mexico, we'll go from 6 lines there today to 4 lines; and then same in Turkey. In both Turkey plants transitioning to larger blades, we actually lose a line in each, so if you do that math, you get to 39, but it's the same number of gigawatts, if you will.
Justin Lars Clare
Right, okay. Okay, that makes sense. And then just on the cost of inspection and the increased focus on quality here, I was wondering. Does that serve as a headwind to your gross margins as you move into Q3 and Q4? How impactful might that be? Maybe you can just speak to that element.
Ryan D. Miller
Justin, this is Ryan. I think a lot of that is behind us. So we went through a bit of a learning curve and some catch-up on some blades that were -- and we talked in the first quarter about how we had -- we slowed down deliveries and so we were catching up on some of those inspection and repairs. We've now got a lot of that behind us. So far this year, and you'll see this disclosed in the 10-Q, we've probably incurred around $10 million of higher inspection and repair than what we had planned at the start of the year. There's probably still a couple million dollars left over the balance of the year, so a total of, call it, $12 million or so, but we think a lot of that increased cost is behind us. I think we've gotten through a lot of the catch-up and learning curve as we've gone through those. It actually has helped us identify areas where we need to go back and build more quality upfront and make it more proactive instead of reactive on the back end and catching it later.
Operator
Your next question comes from Eric Stine from Craig-Hallum.
Eric Andrew Stine
So I'm just trying to think through these quality issues and what this could mean longer term. I mean clearly it sounds like you feel, that one quality issue you're having and the smaller ones, that you've kind of got those ring fenced. And historically I don't think you've had many of these warranty programs. I mean, do you see this at all changing kind of the whole insource-versus-outsource dynamic? Or I mean, is this something that could permanently slow down that move to larger blades, which I guess would mean fewer start-ups and transitions for TPI?
William E. Siwek
Yes, I don't think it changes the outsource-insource discussion, quite frankly. And new product introduction has already slowed down. I mean you just have to listen to what our customers are saying publicly about new product introduction and the need to standardize and industrialize and running these and modularization of the turbine. And so we're seeing that already, but I do not think it impacts the insource versus outsource. If the market goes to where we think it needs -- or where the -- where many think it's going, there's going to have to be a ton more capacity in the market over time. And I don't think the OEMs are going to want to absorb or are going to have the appetite to spend that capital on blade plants, quite frankly, or on other manufacturing facilities, so no. I don't think it impacts it, but I think, to your point, we have not had significant warranty issues in the past. We see this as a unique event for us. And we will learn from it and continue to improve what we do and work very closely with our customers. One of the important things that we've talked about in the past is collaborating more closely and deeper upfront with our customers so that we can minimize the potential for risk areas in the manufacturing process and in turbine design. And our customers recognize that. They are engaging with us more deeply and earlier, and I think that will help in the long run with minimizing the types of risky manufacturing operations that we see at points in time today.
Eric Andrew Stine
Got it. And when you're talking about the 2 lines in Turkey, going from 3 to 2; and then same in Mexico, I mean, is that -- or are those plans that are already underway and -- or kind of are already starting and so they're not necessarily subject to kind of the slowdown you're seeing on the activity side? Or...
William E. Siwek
Yes. So no, it has nothing to do with that. Let's be clear. In Turkey, it was -- we have 2 plants there. 1 is going from 4 lines to 3 lines. 1 is going from 7 lines to 6 lines. And in Mexico, the 1 plant we have running a smaller blade today, when we go to a bigger blade, that goes from 6 lines to 4 lines, producing the same number of gigawatts off of 39 lines as we were off of the 44 that we talked about before, so it's not about a reduction in demand for what we're doing. It's a larger blade. Therefore, you need fewer sets to hit the same number of gigawatts. Now over time, might we expand a couple of those plants to add more lines to it? Absolutely. Or we'll look for additional capacity in the same geographies or new geographies.
Eric Andrew Stine
Got it. All right, that's helpful. Maybe just quick just on the automotive side: You mentioned strategic alternatives. I mean it doesn't sound like that strategic alternative would include selling or monetizing that business. So maybe just some of the thoughts or the more likely forms that might take.
William E. Siwek
No different than we've talked about, Eric. It's we're looking at all options. It could be partnership, joint venture, a separate source of capital, so there's a multitude of different things we're looking at. We're actually looking at probably 3 different forms today, so those discussions are progressing. We hope to be able to give you more information later in the quarter.
Operator
Your next question comes from Joseph Osha from Guggenheim Partners.
Joseph Amil Osha
I just wanted to talk about next year a little bit. You've got this Oaktree pref dividend coming in. It's $40 million. You have not, based on my sort of pass through your model, ever generated that much free cash at least since 2015 or so, so it would seem to me that you're going to be burning cash next year to meet all of your obligations. Is that a fair observation?
Ryan D. Miller
Joe, I think a lot of this, for us -- I think we feel pretty comfortable with our liquidity position today. Could we be burning some? If we do, we believe it would be a very modest amount. A lot of this is going to depend on the start-ups and the timing of those start-ups. It's going to be dependent upon the amount of customer advances that we can get to help fund those start-ups and transitions, but as we're currently thinking about '24, obviously we're not yet ready to guide to '24 yet. But we laid out what we think our sources and kind of uses of cash are next year. And I think the one thing that you may discount a little bit right now is our ability to go generate cash out of our balance sheet. And so you'll see us focused in a lot on that over the balance of this year and next year to go help fund some of that. We do have different areas of inefficiency. And certainly, when you're a -- in a time period we're in today where we've had some delays in deliveries and we're working through some inventory issues, when you're going through that, we'll help -- we'll monetize that as we go through the balance of this year, but we will certainly plan and work to make sure that we can get -- make it through this. And we initially guided, back when we put out our kind of mid- to long-term guidance, that to light back up all of our idle lines, it's probably in the $25 million to $35 million range; still believe it's in that range. And so I kind of contain the CapEx for that -- to that range, probably spending about 1/3 of that this year. And then the rest of it will primarily come in '24.
William E. Siwek
The other thing to think about, Joe, this is the easiest way to think about it, I think, is we eliminate the losses that we're incurring at the -- at our Nordex Matamoros plant. That covers the dividend, just FYI.
Joseph Amil Osha
So say that again. The losses that you are currently incurring from Nordex Matamoros, that, remind me exactly when that goes away.
William E. Siwek
That -- the contract itself ends June 30, 2024, but we've agreed to a bit of a different structure in 2024, so we'll eliminate virtually the entire loss that you would -- that you'll see this year from Matamoros. So that basically pays for the dividend.
Joseph Amil Osha
So that's $40 million right there.
William E. Siwek
Yes.
Ryan D. Miller
Yes, yes. Joe, in the first half of this year -- we'll, you'll see this disclosure in the Q. We've lost about $20 million. Last year, it was about $40 million, so it's been going on about a $10 million run rate pace [a quarter].
William E. Siwek
[Yes]...
Joseph Amil Osha
I'm sorry. Not to pick -- all right, a quarter. All right, that was my question. And just do you guys have at this point any kind of covenants? I assume not on the convert, but is there any kind of covenant on this Oaktree financing in terms of cash flows that you've got to maintain or any of that or...
William E. Siwek
So we've got -- we have to maintain $50 million of cash in the U.S. and we have to get approval for CapEx over $30 million annually.
Ryan D. Miller
Yes, yes. And one other one is we have an $80 million limit on our other debt that we can maintain, but we do have consents obviously from Oaktree for both the purchases of the wind turbine and for the convert.
Operator
Your next question comes from Sherif Elmaghrabi from BTIG.
Sherif Ehab Elmaghrabi
First, on moving to larger turbine production lines, how long does it take to shift production from one blade type to another? Is it a function of winding down supply agreements? And what -- can you remind us what kind of CapEx is associated with that?
William E. Siwek
Yes. So in most cases, if we're doing a transition, it's in an existing facility that's already got the basic CapEx in it. There are occasions when -- if we're going to a much larger blade, where you might have to do a crane upgrade, which might be a couple million dollars. The biggest CapEx, though, is the tooling itself or the mold. And the mold is paid for by our customer, so the CapEx on a transition, short of having to extend a building or change a crane out, is usually relatively small. The time it takes to do it, part of that depends on the part of the world we're in, whether we're working 5 days, 6 days, 7 days a week, but it's generally -- to go from stopping of the old mold, removal of that mold, install of the new mold and then ramping it back to serial production is probably around a quarter.
Sherif Ehab Elmaghrabi
And then -- and the field service business, a bit of a smaller point here, but how long do you expect warranty claims to pull technicians away from doing field service work?
William E. Siwek
I would say a lot of that is happening this year. We'll have a little bit spill over into next year, but I think, probably by the end of the first quarter or so next year, we should be through most of it. Is that...
Ryan D. Miller
Yes, I think, back to a more normalized...
William E. Siwek
Yes, back to a more normalized rate. So it will impact us for the balance of this year and a little bit of next year and then we should be back at kind of a more normalized level of warranty versus revenue-generating work.
Operator
Your next question comes from Pavel Molchanov from Raymond James.
Pavel S. Molchanov
I want to zoom out for a moment about kind of this whole quality control discussion in the wind space. Ever since Siemens started talking about this right around a month ago, are -- do you have a sense that there is almost like a witch hunt to try to find and kind of nitpick things that perhaps would not have been regarded as serious issues until that rather kind of high-profile headline, so an overreaction, so to speak?
William E. Siwek
Pavel, I have to agree with you. I think clearly there are some -- and that's why, when I was talking a little bit about it after the first question -- I mean we're categorizing the issues, right, where it's -- yes, it's something that really needs to be fixed because it could impact the longevity of the blade. Others are, yes, we ought to watch that, but likely it's not going to impact performance or longevity. And then there are other things that, well, yes, you could fix it because it's a nonconformity. It's not going to make it perform any better. It's not going to do anything different. In fact, it may create more of a problem if you fix it, as opposed to leaving it as is. And so I think there is a heightened sensitivity, especially for those customers of our customers who may not completely -- I'm trying to say this in the right way: who may not completely understand how these blades are built and the fact that it is a very manual process. And first-pass yield on a blade is generally 0 because there are always some nonconformities because it is a manual process where you're infusing resin on a blade that's as long as a football field. So you're going to have some of that stuff, so I do think there's probably a heightened sensitivity around nonconformities in the blades that maybe we didn't see in the past.
Pavel S. Molchanov
Okay, let me turn to the electric vehicle. I guess it was 7 years ago. It feels like ancient history that TPI started supplying these bus bodies to Proterra. And since then, we're watching every commercial fleet is electrifying buses, trucks, vans. And I'm curious why carbon fiber has not become kind of the universal or mainstream solution of -- because it feels like it's actually shrinking. What do you think happened there?
William E. Siwek
Yes. Carbon fiber is too expensive, Pavel. That's why, but what we're doing is not carbon fiber. It's selective use of carbon fiber in key stress points that you need to use carbon, but most of what we're doing is more glass fiber, as opposed to carbon fiber. I think, as if -- I mean, if you look at what we're -- I wish I could tell you who we were working with, but if you look at who we're working with and what we're working on: There is some real traction taking place with the composites side of the business from an automotive standpoint, as they recognize it's not just lightweighting but it's the durability. It's the performance, especially in a battery enclosure or especially in certain of the structural panels and large panels, with some of the unique technology we have. So I think you will see it pick up, especially in the commercial space, but again it's glass fiber mostly, not carbon fiber.
Pavel S. Molchanov
Right, okay. I mean point well taken. Are you working with Workhorse still?
William E. Siwek
No. We haven't been working with them for, I don't know, over 2 years now, I think.
Pavel S. Molchanov
Okay. And GM...
William E. Siwek
We have done some work with GM that we've announced publicly. I can't say what we're doing now.
Operator
(Operator Instructions) Your next question comes from Julien Dumoulin-Smith from Bank of America.
William E. Siwek
Julien, you're double dipping on us.
Julien Patrick Dumoulin-Smith
I'm back and you better believe it. I just wanted to follow up, if I can, on really pushing on the impacts, the ongoing impacts, here. I mean I know you said by the end of 1Q that you should be seeing some normalization on sales or at least the field sales, but just can you elaborate a little bit on sort of the ongoing OpEx and the time line there? I mean it sounds like -- again you've reaffirmed, a couple of years out, you see an ability to get back to that high single digits outlook, but I just want to make sure we understand the full extent of the revenue and the OpEx impacts. And then separately, I'll throw the second question in there, on the U.S. expansion. I [thought] there was some talk about looking at the Western U.S. for further facilities. Is Mexico now in lieu of that?
William E. Siwek
So I'll answer, first, on the OpEx and revenue. I would say on the OpEx side, Julien, we've -- as Ryan indicated, we've spent about $10 million in the first half of the year. We'll spend about $2 million in the back half, as we've caught up; and as we've gotten more efficient at kind of the new procedures that we're following, that we've developed, as well as some of the new testing techniques and inspection techniques. I think it'll -- we won't have a increase in OpEx because I really think, over time, we're actually going to be -- we're actually going to find that we can reduce the costs of inspection because of what we'll be doing upfront and how we'll be doing it differently. So I think long term it's actually -- it will be helpful to us. And on the revenue side, again it's -- and from a field service standpoint, we should -- we're still hiring. We're still trying to expand in all regions that we serve today. The limiting factor is hiring [techs], quite frankly, but we're still moving down that path. And we think we'll get back to a more normalized split of revenue-generating versus warranty work by the end of Q1 of '24. And then on the expansion, no, it's not really in lieu of, Julien. It's in addition to, is my belief.
Julien Patrick Dumoulin-Smith
Yes, okay. So that's actually still in play here for -- in the near or medium term and your discussions are ongoing...
William E. Siwek
Yes. Ongoing discussions with multiple parties, yes.
Julien Patrick Dumoulin-Smith
Okay, all right, but is that a this-year thing at this point given the push-out?
William E. Siwek
I'm sorry. Is that -- what...
Unidentified Company Representative
(inaudible) this year...
William E. Siwek
This year...
Julien Patrick Dumoulin-Smith
Is that still [viably an FY '23 issue]?
William E. Siwek
No. I'm -- I don't think it was ever a '23 -- I mean it's -- as far as building a facility, no; as far as identifying. And we're in that process right now, so yes. I mean, is it possible that something got -- gets announced before the end of the year? Certainly. My guess is it's probably more of a late '23 or a '24 time frame.
Julien Patrick Dumoulin-Smith
Got it. Sorry. I didn't mean to press you too much on that. And to clarify an earlier -- Nordex, the payment there offsets the dividend payment, if I heard that right...
William E. Siwek
Yes. The -- yes. What we're losing in that factory today is essentially equivalent to what the dividend payment will be for Oaktree next year.
Julien Patrick Dumoulin-Smith
All right. Or said differently, the liquidity that Nordex is paying you is effectively equivalent.
William E. Siwek
I'm sorry. Say that again...
Julien Patrick Dumoulin-Smith
The Nordex, like, compensation is effectively equivalent at that point.
Ryan D. Miller
Yes, that's correct, Julien. We get back in '24 at a pace that we're pretty close to breakeven. So the losses we experienced last year and this year which were about $40 million a year or about $10 million kind of run rate a quarter, that effectively ends, beginning January 1, 2024.
Operator
Bill, there are no further questions at this time. Please proceed with your closing remarks.
William E. Siwek
Sorry. I had a, took a sip of water. Thanks again for your time today and continued interest and support of TPI. I look forward to speaking with you again shortly and next quarter. Thank you.
Operator
Ladies and gentlemen, this concludes your conference call for today. We thank you for joining, and you may now disconnect your lines. Thank you.
