Good afternoon, everyone, and welcome to First Solar's third quarter 2019 earnings call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. At this time, all participants are in a listen-only mode. As a reminder, today's call is being recorded. I would now like to turn the call over to Adrianna DeFrisco from First Solar Investor Relations. Ms. DeFrisco, you may begin.
Thank you. Good afternoon, everyone, and thank you for joining us. Today, the company issued a press release announcing its third quarter 2019 financial results. A copy of the press release and associated presentation are available on the First Solar website at investor.firstsolar.com. With me today are Mark Widmar, Chief Executive Officer, and Alex Bradley, Chief Financial Officer. Mark will begin by providing a business and technology update. Alex will discuss our financial results for the quarter and provide updated guidance for 2019. Following their remarks, we will open the call for questions. Please note this call will include forward-looking statements that involve risks and uncertainties that could cause actual results to differ materially from management's current expectations. We encourage you to review the safe harbor statements contained in today's press release and presentation for a more complete description.
It is now my pleasure to introduce Mark Widmar, Chief Executive Officer. Mark?
Thanks, Adrianna. Good afternoon, and thank you for joining us today. I'll begin by noting our third quarter earnings of $0.29 per share. We had a variety of sources contributing to our Q3 EPS result. Firstly, the third quarter reflected our strongest quarter of the year with respect to third-party module sales, driven primarily by a significant increase in the volume of watts sold compared to previous quarters. Secondly, on the systems segment side, we successfully closed the sale of our 72-megawatt Seabrook project and are continuing the momentum with the recently signed transactions for the sale of our 150-megawatt Sun Streams, 100-megawatt Sunshine Valley, and 20-megawatt Windhub projects. These deals remain subject to certain conditions precedent for closing, which we expect to be satisfied during the fourth quarter, resulting in a meaningful contribution to our full-year revenue and earnings.
Finally, reflective of our modules' outstanding field performance, lower-than-expected warranty return rates, and other factors, the third-quarter EPS result benefited from a reduction to our product warranty liability reserve. Alex will provide further details on this later in the call. From a module production standpoint, we continue to make significant strides in the ramping of our Series 6 module capacity. As of the beginning of this week, we started the ramp of our second Perrysburg factory approximately three months ahead of schedule. This marks the fifth Series 6 factory where we have commenced operation within a period of only 18 months. Across our manufacturing fleet, we had 4.2 gigawatts of Series 6 nameplate capacity at the end of the third quarter, increasing to 5.5 gigawatts this week with the addition of Perrysburg 2.
From a booking perspective, we added an additional 1.1 gigawatts of net bookings since the last earnings call, bringing total revenue-expected shipments to 12.4 gigawatts. Accordingly, we have significant visibility regarding the module business into 2021. During the quarter, we announced an evolution in our EPC delivery approach, transitioning away from an internal EPC execution model and moving towards leveraging the engineering, procurement, and construction capabilities of innovative partners in the United States. While we are proud of our history as a leader and pioneer in the global PV EPC industry, there were several drivers to this decision. Among these is leveraging the advanced capabilities of our partners, whose primary or sole business focus is EPC execution, which enhances product cost-competitiveness and de-risks project execution for First Solar. Moreover, our ability to execute this transition reflects an improved balance of system compatibility of our Series 6 module.
In the past, one of the drivers of our EPC business was to enable the cost-effective installation of our Series 4 product with its unique form factor. Our Series 6 module, which has a larger form factor consistent with other technologies, de-emphasizes the need for our in-house EPC solution. As previously noted, this transition to leveraging third-party EPC services in the U.S. is not expected to impact any projects under construction and slated for delivery this year. Turning to slide four. As mentioned previously, our second Perrysburg facility commenced production approximately one quarter ahead of schedule, and we are expected to ramp to be completed by the end of the first quarter of 2020.
As discussed on prior earnings calls, by removing certain finishing constraints through inventory buffers and limited additional tooling, we expect to increase Perrysburg 2 nameplate capacity to 1.3 gigawatts, resulting in a total nameplate capacity in Ohio of 1.9 gigawatts. The accelerated start of manufacturing has the potential to add up to 80 megawatts of incremental output in 2019. However, due to production timing and quality review, we do not expect any impact to 2019 revenue guidance associated with this change. These same process improvements will be applied across the balance of the production fleet, and we expect a comparable increase in nameplate capacity once they are fully ramped in 2020. Starting on slide five, I'll provide an update on our Series 6 production metrics.
On a fleet-wide basis since the Q2 earnings call, we continue to see significant operational improvements. Comparing October month-to-date metrics against the month of July, megawatts produced per day is up 8%. Capacity utilization has increased six percentage points to 100%. The production yield is up two percentage points to 93%. The average watts per module has increased four watts, and the top production bin is 435 watts. Finally, the percentage of modules produced with an anti-reflective coating has increased five percentage points to 96%. A combination of our efficiency improvement program and increased ARC utilization has led to a significant improvement in the module bin distribution. The ARC bin distribution between 420-watt to 430-watt modules is up five points to 92% of production. Relative to our efficiency roadmap, there were a couple of noteworthy accomplishments during the quarter.
Firstly, following the implementation of the latest process refinements on our lead line in Ohio, we have started to produce our first 440-watt bin modules. Leveraging these refinements with upcoming process nodes, we have recently provided sample modules to Fraunhofer, who have measured and validated a 19% aperture area efficiency, which is 447 watts peak and is a new world record CdTe module. This is a tremendous accomplishment by our world-class R&D and manufacturing teams and further validates the near-term compatibility of our technology. Secondly, our continued investment to improve long-term performance of CdTe cells has resulted in a significant scientific advancement. As some of you may know, copper serves as a critical role in the CdTe device. Through our recent R&D efforts, we have successfully replaced copper with an alternative material, which dramatically improves device performance.
More importantly, these high-efficiency devices demonstrate an improved long-term degradation rate, a significant benefit to PV plant economics. While we have yet to transition this advancement to our commercial product, we expect the combination of improved efficiency and long-term degradation will further enhance the competitive advantages of our technology and gives us confidence in our long-term roadmap. At the beginning of the year, we laid out an aggressive Series 6 cost per watt reduction target for 2019. As we expected, a drop in Series 6 cost per watt from Q1 to Q4 of approximately 30%. Relative to our expectations through Q3, we are pleased with the progress made and are slightly ahead of the roadmap laid out during the 2018 year-end earnings call, which took place in February. On a fleet-wide basis, throughput is ahead, efficiency is in line, and yield is slightly behind.
By region, our low-cost factories are performing well, and specifically with the faster ramp of our second Vietnam factory, we are seeing a cost per watt benefit. However, our Perrysburg facility is behind expectation and is facing some significant headwinds. We continue to see challenges to the bill of materials, including aluminum and glass, as we work through our supply chain strategy to mitigate the impact of tariffs. In addition, our labor and sales freight costs remain above plan. Looking forward to Q4, we expect continued improvements in our low-cost factories, which will end the year in line with our expectations. However, Perrysburg will be challenged for the reasons mentioned, as well as the earlier production ramp of our second factory.
Our current view is that on a fleet-wide basis, we will exit the year approximately a half a penny higher than the target we set at the beginning of the year. Turning to slide six, I'll next discuss our bookings activity. In total, we have 5.4 GW of net bookings in 2019, including net bookings over 1.1 GW since the last earnings call. After accounting for shipments of 3.8 GW in the first three quarters of the year, our expected future shipments are 12.4 GW. Our net bookings since the last earnings call were almost exclusively for Series 6 product and were approximately 85% in North America, with the remainder in Europe. This includes 0.3 GW previously included in our mid- to late-stage pipeline as signed but not booked opportunities. In these most recent bookings, we are sold out through the remainder of 2019 and full year 2020.
We are largely contracted through the first half of 2021 and approximately two-thirds booked relative to the 2021 supply plan of 6.5 gigawatts. Note, this excludes anticipated future systems projects not currently recognized as bookings. We also now have collectively over one gigawatt booked for 2022 and beyond. Our debookings include a 75 megawatts for our Willow Springs project. This debooking is relative to our previously disclosed petition to the PG&E bankruptcy court to terminate the PPA. During the third quarter, the bankruptcy court granted the petition and we terminated the PPA, which will allow us to re-market the project to another offtaker. With our year-to-date net bookings of 5.4 gigawatts, we have achieved our target of a one-to-one book-to-ship ratio in 2019, approximately two months ahead of year-end and continue to see ongoing demand through the remainder of the year.
Slide seven provides an updated review of our mid- to late-stage bookings opportunities, which now total 8.1 gigawatts DC, an increase of 2.1 gigawatts from the prior quarter. When factoring in the bookings for the quarter, one gigawatt of which were included as opportunities in the prior quarter, our mid to late-stage pipeline increased by 3.1 gigawatts. Additionally, the pipeline includes 0.3 gigawatts of confirmed opportunities awaiting satisfaction of outstanding conditions precedent before being recorded as a booking. As a reminder, our mid to late-stage pipeline reflects those opportunities we feel could book within the next 12 months, and is a subset of a much larger pipeline of opportunities, which includes 15.2 gigawatts DC, which increased 1.9 gigawatts from last quarter.
This includes 1.6 gigawatts of opportunities in 2020, which provides demand resiliency to our near-term production, while maintaining 13.6 gigawatts of demand would be for module deliveries in 2021 and beyond. In terms of geographical breakdown of the mid to late-stage pipeline, North America remains the region with the largest number of opportunities at 5.8 gigawatts. Europe represents two gigawatts, with the remainder in Asia-Pacific. In terms of segment mix, our mid to late-stage pipeline includes approximately two gigawatts of systems opportunities across the United States and Japan, with the remainder module-only sales. Our energy systems business continues to perform strongly with an additional one gigawatt contracted since our previous earnings call. This brings new bookings in 2019 to 2.6 gigawatts, and our total energy services portfolio of assets under contract to nearly 14 gigawatts global.
Before I turn the call over to Alex, I would like to cover a few items. Firstly, I would like to note that we are pleased with the United States Trade Representative's decision earlier this month to withdraw its exclusion for bifacial modules from the Section 201 import tariffs. This decision supports a level playing field for manufacturers such as First Solar that innovate, manufacture, invest, and create jobs in America. Secondly, as First Solar celebrates its 20th year since its founding, we would like to take a moment to reflect on the incredible strides the company has made and that the solar industry has experienced over that time. What was once a producer of a niche technology has evolved into a global company with upstream and downstream capabilities.
From establishing the industry's first PV module recycling program in 2005 to breaking numerous cost per watt barriers, to having over 10 gigawatts of solar assets under operation and maintenance management, to shipping over 25 gigawatts of modules since our founding, and today having 5.5 gigawatts of capacity of our new Series 6 module. First Solar has continued to evolve its business model to remain competitive and differentiated in a constantly evolving market. We have done all of this as a U.S.-headquartered company with its manufacturing and technology roots in Perrysburg, Ohio, and our advanced research and technology capabilities centered in California. Among our competitive differentiators, including our technology differentiation, industry-leading balance sheet strength, and a sustainability advantage, we are in a fortunate position of being sold out through the second quarter of 2021, with significant bookings visibility throughout the balance of that year.
This visibility over a multi-quarter horizon has allowed us to be discerning of the opportunities that have availed themselves over the year. Finally, our competitive and financial position enables First Solar to continuously evaluate the cost structure, competitiveness, and risk-adjusted returns of each of our product offerings, including the module, development, and O&M businesses. As discussed, we have evaluated our EPC capabilities and are transitioning to a third-party execution model. As a result of this transition, approximately 100 associates directly associated with our EPC capabilities will leave the company. This evaluation is not being done in isolation. Since announcing the launch of Series 6, we have contracted over 15 gigawatts and have created a position of strength with a multi-year pipeline. However, we cannot be complacent.
Rather, now is the time to challenge ourselves to secure the right long-term sustainable cost structure for our module manufacturing, development, and O&M businesses in order to best position each for success over the next decade. We expect to conclude of this very in-depth review and communicate the results at the end of the first quarter of 2020. I'll now turn the call over to Alex, who will discuss our third quarter financial results and provide updated guidance for 2019.
Thanks, Mark. Starting on slide nine, I'll cover the income statement highlights for the third quarter. Net sales in Q3 were $547 million, a decrease of $38 million compared to the prior quarter. The decrease was primarily a result of reduced systems project sales in the U.S. and Australia, partially offset by the sale of the Seabrook project in the U.S. and higher module sale volumes. On a segment basis as a percentage of total quarterly net sales, our systems revenue in Q3 was 32% compared to 61% in Q2. Gross margin was 25% in Q3 compared to 13% in Q2.
The systems segment gross margin was -5% and was negatively impacted by several factors, including low overall revenue recognized in the quarter relative to the systems segment fixed costs, a higher mix of lower margin EPC projects relative to self-development projects, and approximately $8 million of charges associated with the decision to transition to a third-party EPC execution model. Our Seabrook asset, which was acquired in late-stage development and was anticipated to have relatively low risk and low development margin compared to our earlier-stage development assets, was impacted by the occurrence of a greater-than-expected variable integration cost under the PPA, which adversely affected the sale value of the project. The module segment gross margin was 40% in the third quarter, which was positively impacted by $80 million reduction in our product warranty liability reserves, partially offset by $6 million of Series 6 ramp-related costs.
The reduction of product warranty liability reserves was driven by an analysis of module return rates, improving module return rates for our most recent series of modules, and updated information regarding our historical module warranty claims experience. The increase in Module Segment gross margin was driven by increased shipments, lower ramp costs, lower cost per watt, and the product warranty liability decrease. The $80 million reduction in the reserve was equivalent to 22 percentage points of Module Segment gross margin. Operating expenses were $197 million in the third quarter, an increase of $11 million compared to Q2. This was driven predominantly by increased startup expense associated with our second Perrysburg factory. We had operating income of $41 million in the third quarter, compared to an operating loss of $9 million in the prior quarter.
This was a result of increased module sales and the product warranty liability reserve reduction, offset by reduced Systems Segment revenue and gross margin, increased operating expenses. We recorded tax expense of $15 million in the third quarter, compared to tax expense of $12 million in Q2. The increase in tax expense for the quarter is attributable to the higher pre-tax income, as well as the jurisdictional mix of income. A combination of the aforementioned items led to third quarter earnings per share of $0.29, compared to a loss per share of $0.18 in the second quarter. We'll now turn to slide 10 to discuss select balance sheet items and summary cash flow information. Our cash and marketable securities balance ended the quarter at $1.6 billion, a decrease of approximately $500 million from the prior quarter.
Total debt at the end of the third quarter is $480 million, almost unchanged from $481 million at the end of Q2. As a reminder, all of our outstanding debt continues to be project-related and will come off our balance sheet when the projects are sold. Our net cash position, which is cash, restricted cash, and marketable securities, less debt, decreased by approximately $500 million to $1.2 billion. Decrease in our net cash balance is primarily related to increased project assets associated with the construction of unsold systems projects and capital investments in Series 6 manufacturing capacity. As of the end of the third quarter, we had approximately $570 million of project assets under construction on the balance sheet. Net working capital in Q3, which includes non-current project assets and excludes cash and marketable securities, increased by $241 million versus the prior quarter.
The change was primarily due to increased accounts receivable and reduced accounts payable. Cash flows used in operations were $318 million in the third quarter. Finally, CapEx was $183 million in the third quarter compared to $179 million in the second quarter as we continued Series 6 capacity expansion. Continuing to slide 11, I'll next discuss updates and related assumptions to our 2019 guidance. Firstly, 2019 costs associated with our decision to transition to a third-party EPC execution model and not previously in our full-year guidance, including asset write-downs and severance costs, are expected to total approximately $10 million. Of this amount, $8 million was recognized in Q3, with $2 million expected to be recognized in Q4. Approximately $1 million are cash charges, with the remainder non-cash, and annual savings from these charges are expected to total $10 million-$15 million.
Secondly, relating to the module business, there are four key updates. The reduction in product warranty liability reserves of $80 million recorded in the third quarter was a non-cash item and was not part of our previous full-year guidance. The earlier-than-previously anticipated startup of our second Perrysburg factory is expected to increase 2019 ramp costs by $10 million, offset by a $15 million reduction in start-up costs. Additionally, Perrysburg 2 production may provide up to an additional 80 megawatts of product in 2019, which is not expected to be sold in 2019. As it relates to our remaining Series 4 production in Malaysia, we're still evaluating the timing of the Series 4 shutdown to facilitate the conversion to Series 6 and any resulting costs. Thirdly, relating to the systems business. In the U.S., we have recently signed agreements to sell our Sunshine Valley, Sun Streams 1, and Windhub assets.
Although still subject to closing conditions, we expect to close and complete the sales in the fourth quarter. Project sale values are in line with our expectations relative to our 2019 guidance, with the exception of an ongoing development item related to one project, which negatively impacted value by approximately $40 million. In Japan, two weeks ago, Typhoon Hagibis passed very close to our Miyagi project. We're thankful there were no injuries to personnel related to the project. However, the impact of heavy rainfall and blocked roads has so far limited our access to the site. The project is in early-stage construction, with the majority of the work performed to date being civil engineering. Until we complete our site assessment, it's unclear what impact this event will have on project construction timing and value.
In addition, in Japan, a recent proposal by the Ministry of Economy, Trade, and Industry, which would levy a new wheeling charge on solar projects with higher feed-in tariff rates, has introduced some uncertainty to the market for equity ownership. This project, Ishikawa and Anamizu, are currently being structured to be sold together using a private fund vehicle. However, we also have the ability to sell these projects as individual assets. We're continuing to evaluate the preferred sale structure and these events in order to optimize the aggregate value of the projects. Whilst the sales of these assets are all included in our 2019 guidance, we may decide on the basis of this evaluation to defer closing until next year, which would push revenue and margin recognition into 2020.
In the event none of these three asset sales close this year, we could see full year 2019 EPS $0.50 below the guidance range, with a corresponding benefit to 2020. Note that our evaluation of the preferred sales structure for our development assets is not unique to these Japan assets. As we've noted in prior earnings calls this year, such as when discussing optimizing the sale of those projects with an SCE offtake, in light of perceived risks related to the PG&E bankruptcy, we will not compromise project value in order to adhere to any particular timetable.
Fourthly, as a result of this uncertainty related to the timing of project sales, our ongoing evaluation of the long-term sustainable cost structure for our module manufacturing development and O&M businesses that Mark discussed earlier, and the wind down of Series 4 production, we decided to push out the schedule for providing our 2020 guidance. Whilst we typically provide our year forward outlook in the fourth quarter of the year, we believe that clarity around these matters and any associated actions will allow us to provide a more informed full year 2020 outlook. We are therefore deferring guidance until the fourth quarter 2019 earnings call, which is anticipated to occur in February of 2020. Lastly, as a reminder, and as previously disclosed, our ongoing class action lawsuit, which was originally filed in 2012, is expected to go to trial in January of 2020.
As we've noted in the past, given the uncertainties of trial, we continue to not be in a position to assess the likelihood of any particular outcome or to estimate the range of potential loss, if any. We continue to believe we have meritorious defenses and are vigorously defending the case. Our guidance does not take into account the financial impact of any resolution of that lawsuit, given the uncertainties of trial. With these facts in mind, we're updating our 2019 guidance as follows. Our net sales guidance remains unchanged. Our gross margin guidance has increased by 50 basis points to a revised range of 19%-20% due to the product warranty liability reserve release, partially offset by reduced margin associated with a specific development issue related to one of our U.S. projects.
Costs related to the transition to a third-party EPC execution model and increased ramp costs associated with the earlier commencement of operations at our second Perrysburg factory. Operating expense forecast has been lowered by $10 million to a revised range of $350 million to $370 million as a result of decreased start-up expense, which is now forecast to be $15 million lower at $40 million to $50 million, partially offset by higher SG&A and variable compensation expense. Operating income guidance has been increased to a revised range of $320 million to $370 million as a result of the above changes. Full year tax expense is now forecast to be approximately $80 million, up from a previous estimate of $70 million due to increased operating income and a shift in the jurisdictional mix of income. Plannings for share net cash, capital expenditures, and shipment guidance are unchanged.
Finally, I'll summarize the key messages from our call today on slide 12. Firstly, we continue to be pleased with the progress of our Series 6 platform, including the significant improvements across key manufacturing metrics and module efficiency. Secondly, with year-to-date bookings of 5.4 gigawatts, we've met our 1-to-1 book to ship target approximately two months ahead of year-end, and we continue to strengthen our contracted position. Finally, we've increased our gross margin and operating income guidance and maintained our full year revenue and EPS guidance ranges. With that, we conclude our prepared remarks and open the call for questions. Operator?
As a reminder, to ask a question, you'll need to press star 1 on your telephone. To withdraw your question, press the pound or hash key. Please stand by while we compile the Q&A roster. Our first question comes from Philip Shen with Roth Capital Partners. Your line is open.
Hey, guys. Thanks for the questions. The first one is around your projects. Alex, I know you gave some detail just now in terms of what to look for in the 4th quarter, but can you specifically comment on your confidence level and potentially, or possibly the risks around closing on time in Q4? Also, is there a Japanese project that is required for 2019? Secondly, as it relates to your shipments, I think in Q3, the implied shipments were 1.6 GW. How many megawatts of that number was recognized in revenue versus megawatts shipped but not recognized in revenue? Also, what was the mix of megawatts shipped that was recognized that was Series 6 versus Series 4? Thanks.
With regards to the projects, in the U.S., we have three assets, Sunshine Valley, Sun Streams, and Windhub projects. We have signed a deal to sell those assets. That was signed after the end of the quarter. You're not seeing that reflected in the financials for Q3. We would expect that to close in Q4. Obviously there's always some risk until you get to close. We're through the commercial negotiation, now it's getting through just final conditions precedent to closing. A high level of confidence in those assets. When it comes to the full year, the forecast has three projects from Japan in it, our Ishikawa asset, Miyagi asset and our Anamizu asset. As of now, those three are all structured to be sold under one fund structure together.
The typhoon that went through Japan has had an impact to one of our assets, and as we mentioned in the prepared remarks, we're just starting to get access to the site and understand the damage there. Now I think from the very preliminary views we have, we think the damage to the site has not been extensive. However, the damage perhaps to the surrounding area and potentially the gentai, we're still waiting to get an assessment of that. When we look through that, obviously there's insurance on the side as well. It's early to say, but our very early indications would seem to be that from an economic perspective, we'll be okay. From a timing perspective, it's still unknown. Those three assets are all in the guidance. They're all due to be sold together.
If we have a delay in one of them, that could potentially delay all of them. As I mentioned in my remarks, we do have the ability to sell those assets individually as well. There could be some delta in value doing so. We've obviously structured the fund because we think that's the optimal structure and optimal value for all three assets. As of now, they're in the guidance. We'll get more information over the coming weeks. The other key is that two things. One, we're not going to compromise the value. If it gets to a point where we could pull those assets out and sell them this year, but we would be destroying value by doing so, we won't do that. We'll keep them all together. The second is, if there is a delay in selling the asset, it's simply a timing shift.
It just pushes out from Q4 into the next quarter. It's not a change in value. It's simply timing of revenue and margin. That's the project piece. Any other shipments?
Phil, I know one of your questions was on the revenue mix. The revenue mix was on about 60% was Series 6 and 40% was Series 4. The other question was around the shipments within the quarter, how many were recognized within the quarter? I just want to make sure I got your question correctly.
I think we calculated 1.6 gigawatts of shipments in Q3. Sometimes you guys ship, but don't necessarily recognize.
Yeah
those megawatts in revenue.
Yeah.
Did you recognize all 1.6? Thanks.
No. You're close to the number around shipments. We actually have a little over 800 megawatts of third-party module sales that have not been recognized. They're deferred revenue at this point in time, effectively, which will be recognized in the fourth quarter. We also shipped another close to a little bit north of 100 megawatts into our own system assets. You've got a combination of about 800 megawatts that didn't get recognized for third-party module sales just based off of shipping terms and trigger for revenue recognition. You got another 100 megawatts or so of Series 6 product that got deferred out into our systems projects, which will be now recognized. A significant portion of that will be recognized associated with our Golden Eye, Sunshine Valley, and Windhub projects that we've signed, which will close in the fourth quarter.
Phil, if you're trying to walk it forward from last quarter, we said last quarter had about 1.4 gigawatts of shipments, and about a third of that recognized revenue, a third of that was shipped to our systems assets, didn't recognize revenue, a third went on to the module business and didn't recognize revenue. As you carry over the module shipments from Q2 that didn't rev rec in Q2, will have rev rec'd in Q3. However, the shipments that went to our systems assets in Q2 may likely not have done still, because a lot of that would have been associated with the U.S. assets that we haven't recognized revenue on yet and completed the sale of.
Your next question comes from Julien Dumoulin-Smith with Bank of America. Your line is open.
Hey, guys. Good afternoon. It's Julien. Just to follow up very quickly on the last question, if you don't mind. Can you clarify just which projects are included in this year's versus next year's net $0.50? Specifically here, with respect to the SCE projects, those are signed but not closed on. Just again, sorry to come back on this. Just want to be extra clear about what's this year versus next year. I'll leave it there. I got to follow up.
Just around the SCE project, I'll let Alex handle the balance of it. Yes, the projects that we just referenced that have now been signed, subject to CPs to close, there's three of them. Those are all with SCE as the counterpart. We're very happy to move those forward. As you know, we've talked about those throughout the other earnings calls earlier in the year. Those are moving forward. That is, again, Sun Streams, Sunshine Valley, and Windhub. Those are all in the process of being signed. They've been signed now and in the process of closing, which we will expect to happen here in the fourth quarter. The other projects that are still moving around are Japan assets, which Alex can talk about.
Yeah, the Japan, the previous answer. Our Ishikawa asset, our Miyagi asset, and our Anamizu asset. All three of those are currently in the guidance forecast for the year.
Got it. Just the $0.50 would be for which one is moving out? Just to clarify that.
The comment given was around if we do not sell those Japan assets, we could see coming in $0.50 below the low end of the range of guidance.
Okay. Just Japan. All right, excellent. If I can continue here. You commented about this $0.005 improvement versus the initial 4Q guidance on S6. How does that position you with respect to further cost reduction edging into 2020 and onwards? How do you think about that roadmap? Certainly good success across the board here. Can you comment more broadly, if you can, especially as you think about the optimization elements?
Again, it's a half penny actually higher than our target, right? If you look at across the fleet, where we're exiting the year will be about a half a penny higher across the fleet. Again, put it in perspective, that was a 30% reduction from where we started the year, and we're talking about a few percentage points that we will be above that number from our current view. We're better than where we had anticipated through three quarters. We've got a couple of headwinds that I mentioned in the call, mainly impacting our Ohio factories. I say factories because when we began the year, we did not anticipate the startup of Perrysburg 2. The target that we gave you would not have had Perrysburg 2 in that number.
Perrysburg 2 startup and again, underutilization and just general ramp and initial production bins being lower than average. While we are now starting using ARC this week. We started up on Monday. ARC will start to be commissioned this week. We actually are getting good bins without ARC, and when we include ARC, we'll see even better bins. They will be slightly lower bins than the overall distribution across the fleet. You got a little bit of startup underutilization, we refer to it as that. Bin distribution is not nearly where the fleet average is, so it creates a little bit of headwinds. Combine that with the tariff-related costs that we have on our U.S. manufacturing, it's impacting the overall fleet by about a half a penny.
We're in the process, we've largely resourced our cover glass to avoid tariffs, but even there, we're seeing slightly higher costs than we would have otherwise. We're still working through our aluminum frame around resourcing of that aluminum that would bring the tariff down. That'll help drive costs down as we go into next year. The other levers around costs will be further throughput advantages. What we've indicated, we already are starting up Perrysburg 2 with higher nameplate capacity, almost another 100 MW. We'll leverage that across the existing fleet so that incremental throughput will drive better cost per watt average. The efficiencies are already improving. Our top bins right now are 435. We just started to reach 440. As we roll that higher efficiency, higher watts across the fleet next year, we're going to see benefits from that as well.
There's a lot of positives that will help drive costs down further next year. We do got to figure out, we still have some labor inefficiencies that we're working through. Sales freight in the U.S., in particular, for various reasons, is higher, and we got to find a way to bring that down. There are some headwinds. As I look across the long-term horizon and ultimate destination of where we'll end up, we're very confident that we'll meet the targets that we set out for with Series 6. We are dealing with some near-term headwinds, and we're evaluating, obviously, what that potential impact, if any, will flow into next year.
Our next question comes from Brian Lee with Goldman Sachs. Your line is open.
Hey, guys. Thanks for taking the questions. First one, just on the gross margin, Alex. Lots of moving parts here for the module segment. Excluding the warranty item, thanks for breaking that out. The gross margin, I guess, is implied to be 18% there. There's some other moving parts. You mentioned the pull forward on Perrysburg 2, I think $10 million of some ramp-up cost and then some other items. If we adjust for all of that, I'm getting to something in the low 20s, sort of as a clean gross margin for the module segment. Is that fair or are we maybe missing something else there? Maybe if you could walk us through some of the moving parts, and then I had a follow-up.
Yeah. You're doing the right math on the product warranty release comes out of COGS out of the module segment. The roughly $40 million reduction to COGS coming out of one of the development items that comes out of the system segment. That's in the full-year guidance, so you're not seeing that in the quarter. In the quarter, you're going to see about $10 million of ramp that's out of the module side, and then about $10 million associated with EPC. If you walk that math, you get 80 minus 40 minus 10, minus 10, getting you to about a 20 delta in gross margin. That's what you see in the half point increase to the guidance.
Okay, fair enough. Just on that point, there's nothing from the Q3 quarterly result that would repeat in Q4? Just to make sure I get my following in. Just on the capacity here. Last December on the update call, you talked about 5 to 5.5 gigawatts of production in 2020 as the target range. I think with Perrysburg 2 pulling forward 3 months, maybe some optionality to keep Series 4 online through Q1 for safe harbor. How should we be thinking about that range? It seems like it has some upside at this point and also with the debottlenecking. Thanks, guys.
The last thing on the quarter. The severance we talked about on the EPC charges associated with exiting the U.S. EPC business. That's $10 million, $8 million hits the quarter. You will see $2 million flow through into Q4. Outside of that, the warranty reserve release was a one-time in Q3, and the development item will hit all in Q4 as we roll through the revenue recognition when those projects are solved. Yeah, it relates to capacity next year for Brian. We're not in a position yet to give guidance on that, but I will say that the early start of Perrysburg effectively allows us now, we'll have more production out of Perrysburg 2 in 2020 than we would have otherwise. I'm really happy with where we're starting off.
The initial bins that we're seeing is coming across the production line is really good. Throughput, we expect to make meaningful progress of ramping up throughput between now and the end of the year. Our leaping off point exiting this year and entering into next year will allow us an opportunity to get incremental production in 2020 out of our Perrysburg 2 factories. The other items in terms of incremental throughput, debottlenecking, other things that we can do across the fleet, that exist as well. We've highlighted that on name plate capacity, we're right now at 5.4. That does not include the additional debottlenecking and other things that we can do across the other factories, so there's a little bit of upside there.
I would say we're comfortable that for sure we'll be at the high end of the range that we previously guided you for production for next year, we'll see where we can go above and beyond that. Clearly, that's one of the things that we were highly motivated to start up production, not only because we were ready, but obviously driving that through and ramping that factory gives us incremental capacity next year and a very strong demand opportunity set for us. Hopefully with that incremental throughput, we'll drive better revenue for next year as well.
Our next question comes from Ben Kallo with Baird. Your line is open.
Hey, thanks guys. Maybe going back to the Analyst Day and your margin targets, could you just talk about where you are with hitting those targets and then how, I guess, changing from the EPC business to module sales changes that overall model. Maybe just a little more context what a half penny within your target means. You can still get to that greater than 20% gross margin on Series 6? Thanks.
Yeah, I guess the way I would look at it, Ben, is margin's obviously a combination of two things. The margin's a combination of what are we able to capture in the market from an ASP, and then what's the actual cost of producing the product. If you look at our bookings, the 1.1 gigawatts that we highlighted, again, with volumes going out not only in 2021 but in a portion of that beyond 2021. The ASPs, when you look across that, are all very good. They're very much in line with, on a blended basis, if you look at the Q that'll come out tomorrow, I think the numbers will apply something around $0.34, somewhere in that range. These additional bookings post the end of quarter will be consistent with that type of number. When you look at that result, a good ASP.
When I also look at the bends we're producing, not only that over time, the energy advantage because, the subtlety that we wanted to highlight in the scientific improvement that we made or discovery that we've made around replacing copper, that'll significantly improve long-term degradation of the product. What we've always said, two things, is not only efficiency, but energy. Whatever we do, it needs to drive better energy. That's a meaningful driver to improve the energy, and that drives value, increased value back into the module and the market price that we can get. When I look at that, I feel confident there. The cost right now across the fleet, the fact that our low-cost regions are where they need to be, that's extremely encouraging. We've just got some challenges we're dealing with in Ohio specifically.
We'll address some of that with, throughput is a great way to help dilute your fixed cost and drive your cost per watt down and highlighting that benefit in our factory, our second factory in Ohio will help. We've got some work, though, to do on the labor, and we've got to find a way to get some labor costs out of Ohio. It's running higher than we had indicated. We indicated that earlier in the year. We've got to work through that. We're just not quite there yet. When I look at the long-term destination of where we will be relative with when I look at where ASPs are and what the cost per watt profile is, we're very comfortable with the original gross margin targets that we set for Series 6. Nothing's changed from that standpoint. This half penny's not going to move that.
I would argue that, again, that $0.005 is being more than compensated by the strong ASPs that we're seeing.
Ben, just to add on the EPC piece. I think we guided that, I want to say about a year and a half ago, nearly two years ago now, to a 5%-10% gross margin business in the EPC. That is going to come out as we don't perform in EPC in-house. We're also, at the same time, as we commented, taking cost structure out, and we pointed to $10 million-$15 million of annual recurring savings as we exit the EPC business. Going forward, you may see a change in top-line revenue as we exit the internal performance of that function. On an overall consolidated basis, given that EPC was a low percentage margin business, you should see a beneficial impact across the consolidated gross margin.
Our next question comes from the line of Jeff Osborne with Cowen and Company. Your line is open.
Good afternoon, guys. I was just wondering, in terms of the post the bifacial decision, can you talk about what you're seeing in terms of the broader market discussions you're having into the closing months here of safe harboring? Is there any heightened activity or is it too late to safe harbor at this point?
Look, we've seen some additional activity related to that. Part of the problem is we don't have a lot of supply. That's been one of the challenges. People are looking at Series 4. We've still got a few opportunities potentially that we could enable some safe harboring around Series 4. That could happen. Look, even though obviously the exemption was there for a period of time, the other thing we'd like to continue to emphasize, as we said on the call, we're very happy with USTR's decision around eliminating the exemption that was provided. We were very happy though, still with the ability to go through and sell through that.
I think when the announcement first came out in June of last year, shortly thereafter, we highlighted that we booked one of our largest orders ever with a single customer, which is about 1.7 gigawatts, and that was the backdrop of that exemption being provided. We feel very confident with our capability of our technology to compete against bifacial modules. There could be certain applications and certain geographies that we could be less advantaged in or maybe disadvantaged. That can happen. We understand the bifacial technology, the capability relative to our technology. We feel confident with our relative competitive position. I also continue to look forward with what we're doing, not only in efficiency, but other ways that we can drive energy performance out of our technology that will further enhance that competitive position against bifacial.
Our teams are working very hard, and we're very aware of the competing technologies that are out there, and we're doing everything we can to make sure that we, again, have differentiated technology and advantage technology. Yeah, there'll be a little bit of opportunity here maybe throughout the year-end, but it hasn't been a real mover for us. It's because we don't have the supply available.
Your next question comes from Paul Coster with JP Morgan. Your line is open.
Yeah. Hi, this is Mark Strouse on for Paul. Thank you very much for taking my questions. Most of them have actually been answered. Mark, wanted to ask you a high-level question. You mentioned the 19% efficiency and the 440-watt panels. As you continue to improve your product, should investors think of that as kind of relatively modest tweaks to Series 6 technology? I guess, how much runway do you have until a Series 7 platform would be potentially required?
Look, the way I'd somewhat describe it is that, at core, we're a technology and manufacturing company, and we have to stay ahead of the game on the technology. There's still plenty of runway to go on Series 6. Again, driving efficiency, improving the energy profile of the product, not only long-term degradation, look at ways that we can improve temperature coefficient, look at ways to the extent we can to improve spectral response, right? Anything we can do to generate more energy, in real-world conditions, because the point I always like to make is that the efficiency that flash tests are labeled on a module is at 25 degrees Celsius. The real-world operating parameters of the module have many other issues to deal with, not only with temperature, but moisture in the air and other things that can adversely impact the performance of the module.
You have the long-term degradation impact, and more that you can do to improve that over time all enhances value of the technology. We have many levers still to go around Series 6, but I'm not going to rest without thinking through. The team's doing a great job to think about, you used the word Series 7, whatever you want to call the next evolution of the technology, well, that's top of mind for us. One of the nice things that we have always had is we've had our advanced research team in California that's always been out in front of the game and looking at other potential technologies and capabilities. We've made investments over the years in early investments in CIGS. We've made investments with monocrystalline N-type to understand that technology. We've made investments to study perovskite.
We're in front of the game, and we're always looking at what can we do to take our technology to the next level. We are looking at the next-generation technology. We have to. In the interim, we're very confident and comfortable with the product that we have and its capability to compete and to further enhance. Is there something more disruptive that could be out in the horizon? There may be. We'll keep working on that and hopefully crystallize those thoughts and communicate when we're ready.
Our next question comes from Michael Weinstein with Credit Suisse. Your line is open.
Hi. Most of my question has been answered, too. Maybe you could just talk a little bit more about where you stand relative to the 40% target below 2016 levels for Series 6 cost reduction, maybe more specifically around that 40% number, where you stand now.
What I would say, that 40% number, if I look at it at, let's say our lowest cost factory today, not only because of just the region, but also because we're running two production lines in Vietnam. Vietnam right now, because of that, and one of the things that will happen when we put our second factory in Malaysia, you'll be able to leverage the tool set across both factories, which we know helps drive throughput and manage through downtime and other things that create headwinds around your cost profile. Vietnam is our advantage factory today. When I look at where that cost is as we exit this year relative to that 40% target, we'll be in a very good position to have achieved that number.
When I look at the fleet, we're slightly behind, mainly for the reasons that I mentioned, and that's what we have to work on in terms of getting at the end of this year. It doesn't mean that the destination is going to change. There's many levers we can get to get to the destination. At this leaping off point, we're about across the fleet, about a half a penny higher than where we need to be, and we've got to figure that one out.
Is this the main reason why you're pushing out guidance into early next year for 2020?
No, Michael, if you look at guidance, the comments are in the script.
Module side is a lot of moving pieces in terms of when we shut down our Series 4 and transition that over to Series 6. If you go back to early 2018, at that point, we'd have said we were planning on shutting down that Series 4 at the end of 2018. There was the surge of demand around the safe harbor especially. We had a large customer order come in that pushed us with that volume all the way out through the back end of 2019. We're actually in a position right now where one of our customers around Series 4 is potentially in some financial distress. Now we have security against those bookings, but we're looking to optimize our risk profile. It may make sense for us from a risk and value timing to potentially reduce some of that Series 4 earlier than planned.
Helps us convert earlier to Series 6 and get to our more advantaged products and technology. There's some uncertainty for us around the module business, around when we make that call. There's significant uncertainty around timing in the systems business. As I mentioned earlier, this is timing, not value. Right now we've got U.S. assets that even since the end of the quarter, we've signed a deal on, but we still haven't closed there. That's a positive step and we'd expect to close that. Those need to close. As commented earlier, we've got significant uncertainty around what's happened in Japan. Only two weeks ago, we saw a typhoon go through, which is completely unforecast and unplanned. We've got all those moving pieces.
I think as Mark mentioned in his prepared remarks as well, we're also doing a deep dive into the cost structure across the entire business. We've made the decision around transitioning our EPC approach to third parties. As part of that, we're looking at the isolated costs that remain in the development business, as well as looking at all of the business units, including the module, to make sure we're cost competitive across every business unit. With all those moving pieces, it's not a case of having one binary outcome. We could guide to an A or a B scenario. In this case, we've got a lot of different moving pieces. It makes it very hard to give an accurate guide. We'd rather get through all of those and come out in February with some more accurate numbers.
Yeah. Along those lines, just to put it back in perspective, because it's been a while. If you look at the last three factories which we've commissioned, the two in Vietnam and then our Ohio factory. While we had a facility in Vietnam, we didn't have any associates. We built a brand-new second facility in Vietnam, and then we built a brand-new facility in Ohio. If you go back and remember what happened when we did the wind down for Kulim in our first factory in Malaysia, there were reductions, and there were decommissioning costs. Reduction because you have to remember that effectively on an equivalent basis, there's about 50% less labor in Series 6 versus Series 4, which is one reason why Series 6 is an advantaged product.
Whatever the timing is of that decision to start the conversion of our second factory in Malaysia, there will be costs associated with that. One-time upfront costs, which include decommissioning and some of them that impact to headcount. If that happens this year, if the wind down plus, as you know there's respective notice periods, we have charges. If it doesn't happen until next year, we don't have charges, right? If we run the factory through to Q1, we have more volume, right? We're just not in a position right now, even at the Series 4 level, to have certainty, because we're working through some of the issues that Alex referenced to say for certain we're going to give you guidance that we know as of now, that reflects our best understanding for 2020.
I don't want to do that in December and then come out in February and have to change it because of all these moving pieces. We feel the best thing to do is to wait until February where there's clarity, and then we can give the best information versus having to guide to something and then revise it within a matter of a couple of months.
Our final question will come from the line of Joseph Osha with JMP Securities. Your line is open.
Hey, I made it. Thank you. Just following on from that question a little bit. If I look at your slide from the end of 2018, the idea, and you alluded earlier, Mark, to six and a half gigawatts of capacity going into 2021. That number would seem to be a little conservative. You're suggesting that you've got five 1.2-gigawatt plants plus Ohio. That's 6.6 right there, and that's before any of these efficiencies that you talked about. All other things being equal, I'm just wondering whether we shouldn't be thinking about north of seven gigawatts of output in 2021.
Look, the way we always try to phrase it is that we are very much demand-driven, right? Demand-driven drives supply in our supply capacity plan roadmap. As we get more and more comfortable and confident with the volume in 2021, again, we're about two-thirds right now, plus you've got to incorporate call it a gigawatt or so of our systems business relative to that original supply plan of six and a half. We're starting to get to a point to have more confidence in 2021. I also like that we're starting to fill up some 2022 and beyond. The better we do with bookings and the more we book out into 2021 and beyond, it gives us a higher level of confidence to optimize and make sure we drive the highest supply plan.
It's always going to be aligned to the demand that we see, highly predictable demand, to ensure we can scale and ramp that capacity efficiently because of the customer offtake requirements for that product. Yeah, we're not in a position on this call to give revised numbers. What we've been measuring ourselves against is what we communicated, and that was a 6.5 GW supply plan for 2021. As we learn more, we'll give you the best information we can at that point in time.
This concludes today's conference call. You may now disconnect.