Good afternoon, everyone, and welcome to First Solar's third quarter 2020 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'd now like to turn the call over to Mitch Ennis from First Solar Investor Relations. Mr. Ennis, you may begin.
Thank you. Good afternoon, everyone, thanks for joining us. Today, the company issued a press release announcing its third quarter 2020 financial results. A copy of the press release and associated presentation are available on First Solar's 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 technology update. Alex will discuss our financial results for the quarter and provide updated guidance for 2020. Following their remarks, we 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, including, among other risks and uncertainties, the severity and duration of the effects of the COVID-19 pandemic.
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?
Thank you, Mitch. Good afternoon, and thank you for joining us today. I would like to start by thanking the First Solar team for delivering a solid third quarter. Our operational and financial results were strong, and market demand for our Series 6 technology continues to be robust. We had a number of highlights since our last earnings call, including record Series 6 quarterly production of 1.5 GW, solid bookings of 1.6 GW, commercial production of a 445 W module, and earnings per share of $1.45, bringing our year-to-date earnings to $2.65. Our Q3 financial results were driven by a module segment gross margin increase as well as sales of systems projects. While significant uncertainty remain as a result of the COVID-19 pandemic, we are pleased with our year-to-date performance.
As a result of the improved visibility provided by the closing of certain systems project sales, we are reinstating financial guidance for the fourth quarter of 2020. Alex will discuss our financial performance and guidance in greater detail. Turning to slide three, I'll first discuss our module segment performance. Year to date, we have produced 4.9 GW, including 4.7 GW of Series 6, with each factory averaging over 100% capacity utilization during the third quarter. Throughput was led by our international factories, which averaged 118% and 119% capacity utilization September and October month to date. Domestically, our Ohio One and Ohio Two factories are performing well, averaging 109% and 121% over the same periods. On a fleet-wide basis, in September and October month to date, megawatts produced per day was 16.9 MW and 17.9 MW. Manufacturing yield was 96.6% and 97.2%.
Average watts per module was 436 W and 438 W the arc bin distribution from 435 W to 445 W modules was 92% and 96%. At the time of our third quarter 2019 earnings call, we had recently validated a new world record 447 W CdTe module. Building on this, we have implemented these learnings and started commercial production of a 445 W module. Continuing this momentum over the next few quarters, we expect our top bin to increase further to 455 W. In September and October month to date, our Vietnam factories achieved a manufacturing yield of 98%. We continue to implement the learnings and best practices across the fleet, with a fleet-wide yield target of 98% for our current manufacturing footprint by the end of 2021. In the future, we see the potential to incrementally improve beyond this target.
As noted previously, continued throughput, module watts, and manufacturing yield improvements will help drive down our module costs a lot. Since the previous earnings call, we have not experienced significant disruption to our manufacturing operations from the pandemic. Much of our ability to mitigate the potential impact stems from our vertically integrated manufacturing process, diversified supply chain, and differentiated CdTe technology. By contrast, the largest PV module manufacturers globally produce crystalline silicon modules using a batch process technology with multiple process steps. None of these manufacturers are fully vertically integrated and rely to varying degrees on third-party sourcing of polysilicon, ingots, wafers, and cells. Productions of a single crystalline silicon module requires each of these process steps, several factories, and multiple days. During the third quarter, several polysilicon producers experienced significant disruptions that hindered their ability to maintain manufacturing operations.
This disruption, coupled with a supply chain largely concentrated among a few Chinese companies, reduced the available supply of polysilicon. The polysilicon price increase that followed resulted in downstream pricing pressures for wafer cells and modules, and consequently for project developers. While the market for polysilicon has since improved, these events highlight the benefits of our vertically integrated manufacturing process, which enables price and delivery time certainty for customer orders within our contracted backlog. From a shipping and logistics perspective, the most significant impact to date remains the challenging global freight market. While limited freight capacity has increased spot rates, our logistics strategy, which primarily relies on forward shipping contracts, has helped reduce this impact. Regarding our capacity roadmap, we haven't received the major equipment required to commence commercial production at our sixth Series 6 factory in Malaysia.
However, as highlighted during our second quarter earnings call, third-party equipment vendors, as well as our U.S.-based associates, are needed on site for tool installation. Currently, all non-citizens traveling to Malaysia must have explicit written permission from the Malaysian authorities prior to arrival and are subject to a mandatory 14-day quarantine period. While several vendors have received the necessary travel approvals, we are continuing to collaborate with the relevant agencies to gain approval for the remaining travel in a safe and timely manner. Delays in the approval process and compliance with required isolation procedures have the potential to impact the timing of commercial production, and consequently, our full year 2021 production plan. Despite this uncertainty, we continue to evaluate opportunities across our existing manufacturing footprint to further increase our production and capacity entitlement.
Touching on the systems segment, our EPS was also favorably impacted by the sale of three projects in Japan and two in India. With the sale of our American Kings project in Q2, the Japan and India sales in Q3, and with the potential sale of the Sun Streams 2 project in Q4 2020, we have a viable path to close each project sale contemplated in our original 2020 guidance from February. Turning to slide four, I'd like to highlight the bookings and shipment activity for the quarter. In September, we were awarded the PPA for 180 MW AC solar project, with the option for future energy storage located in Arkansas. This project will support the clean energy needs of three General Motors facilities in the Midwest starting in 2023.
Building off of this and the recent PPAs we have signed with Verizon and Dow, we are witnessing leading corporations taking bold steps to reduce their environmental footprints, and doing so supported by technology developed and manufactured in the United States. As the only U.S.-headquartered company among the 10 largest PV module manufacturers globally, with a differentiated CdTe technology yielding the lowest carbon footprint and water usage and a leading PV module recycling program that recovers 90% or more of the glass, metals, and CdTe semiconductor materials, we are well positioned to address this market need. Additionally, it has been an active quarter for our systems business in Japan as we continue to adding to our contracted backlog with the addition of two projects totaling approximately 80 MW. From a third-party module sales perspective, Series 6 demand has been robust.
Among other bookings, as announced last week, we secured 0.9 GW of volume from Vistra Energy for delivery scheduled in 2021 and 2022. As part of this deal, our Series 6 technology will support six projects in Texas, a region that leverages our temperature coefficient, spectral response, durability, and quality advantages. As a U.S. solar technology provider, we are proud to play a supporting role in Vistra's commitment to achieving net zero carbon emissions by 2050. As highlighted during our Q2 earnings call, we've had a significant volume of 2021 opportunities that were in late-stage negotiations but were delayed due to uncertainty in the tax equity market. While Alex will provide a more detailed tax equity market update, I would like to note that visibility into 2021 tax capacity has modestly improved. We secured 0.5 GW of 2021 opportunities since the previous earnings call.
Additionally, demand in 2022 and 2023 have been strong with 0.9 GW of bookings since the previous earnings call. As a result of the recent systems and third-party module wins, net bookings since the previous earnings call totaled 1.6 GW across 1.5 GW of third-party modules and 0.1 GW of systems bookings. Additionally, while not yet meeting all the requirements of a booking, we have contracted 0.6 GW subjects to conditions precedent for expected deliveries in 2021 and 2023. Note, the project associated with General Motors PPA has been sold in conjunction with a module purchase order and has been recognized as third-party module booking. Including these most recent bookings, we have 6.7 GW booked for deliveries in 2021 and 3.6 GW booked for deliveries across 2022 and 2023. Q3, we shipped 1.2 GW resulting in year-to-date shipments through the end of the third quarter of 3.7 GW.
As mentioned during our Q1 earnings call in May, our shipment profile has been back weighted to the second half of the year. Despite this profile, our year-to-date shipments, including the third quarter, have been below our expectations from the start of the year, largely due to the accommodations for COVID-19 driven customer project and financing delays. Before delving into the specifics of our pipeline of bookings opportunities, it is important to highlight that some of the trends we are seeing, including the impact of COVID-19 on the near and long-term growth of solar installations globally. In the United States, the EIA forecasts that approximately 14 GW of utility-scale solar capacity will be added in 2020. This strong demand is led by several states including Texas, California, North Carolina, Nevada, and Virginia, each with near-term development pipelines exceeding 1 GW .
The continued growth of utility-scale solar despite the pandemic-related headwinds speaks to the relative health of the U.S. market. Internationally, the impact of the pandemic has varied by market. While China remains the world's largest solar market, and installed capacity is expected to increase year-over-year, it has seen project completion timelines slip due to the pandemic. Despite these challenges, the country's 14th Five-Year Plan, scheduled to be launched in 2021, is expected to call for targets of at least 60 GW per year of installed PV capacity, or approximately 300 GW over the duration of the plan. In Europe, we anticipate a contraction in new installed capacity as countries like France extend project COD deadlines by six months to accommodate for COVID-19-related delays.
In India, despite a five-month COD extension, delays caused by a combination of the pandemic and the seasonal monsoons are expected to take a toll on the country's aggregate installed capacity this year. While the global PV industry has clearly not been immune to the pandemic's impact, some developments this year will shape the long-term future of the industry. The first of these is a range of new policies designed to decarbonize electricity and mobility, further while powering post-pandemic economic recovery plans. Arguably, the most wide-ranging example is the European Green Deal, which is aimed to transform the bloc into a carbon-neutral economy by 2050, by decarbonizing electricity and transportation. The Green Deal, which could make solar the number one source of electricity in Europe by 2025, is an example of how political leaders are bundling post-pandemic economic recovery with decarbonization commitments.
The other comment I would like to note is the growing recognition of the importance of self-reliance and a diversified solar supply chain in some of the world's biggest solar markets. A combination of factors, including governmental policy, increasingly tense bilateral relationships, the pandemic, and pricing and supply volatility in the crystalline silicon industry, has reignited the debate around risks posed by allowing a single country to dominate the PV solar supply chain. Responses have been varied, with new rules that favor PV modules with a lower carbon footprint in South Korea, while India and Europe have renewed talks on domestic manufacturing. Earlier this month, the United States, the President issued a proclamation revoking the exemption of bifacial panels from the application of the Section 201 safeguard tariffs.
Although this exemption is currently subject to a temporary restraining order preventing the presidential bifacial exemption revocation from taking effect, the common thread, however, is an underlying desire to boost supply certainty and security while safeguarding domestic manufacturing from unfair competition. In summary, we believe our investment thesis remains inviting as we are well positioned to benefit from the current dynamics in the solar industry. As shown on slide five, our mid to late-stage pipeline of opportunities remains robust and has increased 0.5 GW despite bookings of 1.6 GW since the prior earnings call. In terms of segment mix, this opportunity pipeline of 8.3 GW includes approximately 7.7 GW of potential module sales, with the remaining representing potential systems business opportunities. In terms of geographical breakdown, North America remains the region with the largest number of opportunities at 7.1 GW. Europe represents 0.9 GW, with the remainder in Asia-Pacific.
As a reminder, our mid to late-stage pipeline reflects those opportunities we believe could book within the next 12 months and is a subset of a much larger pipeline of opportunities which total 16 GW of opportunities in 2022 and beyond. From a cost perspective, we indicated during our Q2 earnings call that our Vietnam factory, we have achieved a 40% reduction relative to our 2016 Series 4 cost per watt. Building on this momentum, as a reflection of our manufacturing execution, we have also achieved this milestone at our Malaysia factory during the quarter. Note, as a reminder, our cost per watt metric includes sales freight and warranty. From a bill of materials perspective, growing solar demand and the emergence of bifacial modules, which generally are dual glass, have contributed to pressures on the supply and cost of PV glass.
Similar to our shipment strategy, our glass procurement strategy largely relies on forward contracts, which have substantially mitigated this impact to date. From a fleet-wide perspective, as a result of our continued manufacturing execution, we remain on track to achieve and potentially exceed our 10% cost per watt reduction target between where we ended 2019 and expect to end 2020. In Ohio, our third quarter core cost per watt produced continued to be higher than our international average. Our U.S. manufacturing provides strategic benefits, and over time, we anticipate a reduction in the cost per watt through the following initiatives. Firstly, by installing additional tools and optimizing the two Ohio factories into one consolidated platform, we expect to increase name plate capacity slightly more than 25% to 2.4 GW by the end of 2021.
With this additional capacity, we are able to amortize the fixed cost structure, including labor and depreciation, over more watts produced. We are starting to see this benefit as reflected in our October capacity utilization. Secondly, as previously disclosed, we have contracted a float glass supplier agreement with a producer in Ohio. We anticipate starting to receive the initial benefits of this agreement in Q4 and continuing into early 2021, with an expected reduction in the associated variable bill of material costs. Finally, our manufacturing yield in Ohio was approximately two percentage points below the fleet average. Through the implementation of learnings from our international factories, we see a path to achieve similar yields at our Ohio factories. Through the implementation of these key initiatives, among others, we anticipate our Ohio cost per watt premium over time to reduce to $0.02, including sales freight.
Turning to slide six, I would like to discuss the relative performance of our technology in the lab versus real-world operating conditions. PV module lab testing protocols were developed in the early days of solar, using standard test conditions of 25 degrees Celsius at a terrestrial standard spectrum. PV modules in the field, however, are exposed to variable conditions, including heat, humidity, dust, and extreme weather events such as wind and hail. Each of these factors cause deviation from lab performance, with the effects varying by technology. Ultimately, life cycle energy produced in the field is what drives project economics. By analyzing the factors that cause divergence from laboratory performance, we can better understand the value proposition of our CdTe technology. Firstly, as it relates to temperature, module device operating conditions can exceed 70 degrees Celsius.
Module wattage, however, is assigned at a lab standard test condition of 25 degrees, and as panels heat up over the course of the day, beyond this threshold, there's a corresponding decline in power. Series 6 has a temperature coefficient advantage relative to crystalline silicon, which is anticipated to increase further with our copper replacement module, meaning CdTe responds more efficiently than crystalline silicon to real-world temperatures. Secondly, due to the unique spectrum of light Series 6 captures, our technology outperforms crystalline silicon on a watt-for-watt basis in humid environments. Thirdly, estimated useful life of PV power plants can exceed 30 years, and as a result, degradation is an important driver of project economics. With the expected implementation of our copper replacement program, we anticipate a reduction in long-term degradation beyond our current warranty of 50 basis points per year.
We expect this innovation will enhance our competitive advantage by increasing life cycle energy and project value for our customers. Finally, as it relates to bifacial technology, while there is a potential for backside energy gain, the ground reflectivity, known as albedo, varies by geography, climate, and season, and is often inversely correlated with hot humid climates. Slide six depicts the relative life cycle kilowatt hours expected to be produced by our equal watts of our copper replaced Series 6 modules, which we call Series 6 CuRe, relative to leading crystalline silicon bifacial modules. As a result of the aforementioned advantages, as compared to leading crystalline silicon bifacial modules, we estimate that our Series 6 CuRe module can produce up to 10% more life cycle kilowatt hours per kilowatt installed in climates with extreme heat and humidity, including Brazil, Central Africa, Southeast Asia, India, and southern United States.
Importantly, when implemented, our CuRe product is expected to be well-positioned in other key markets with more moderate climates, including France, Spain, Japan, and the Midwestern United States. We expect to begin delivering Series 6 CuRe modules in the second half of 2021. Turning to slide seven, I would like to review a framework that highlights the factors that influence ASPs. Starting with bifacial, while backside energy gain is accretive to ASP with only a modest increase to manufacturing cost per watt, the downstream costs related to additional balance-of-system structures, increased vegetation management, and higher cost of capital associated with albedo uncertainty, are partially offset to this ASP benefit. As it relates to crystalline silicon with larger form factors, the potential ASP benefit largely stems from the dilution in the manufacturer's fixed bill of material costs rather than an increase in energy density.
This potential cost reduction, which may be passed through to the customer, is partially offset by the downstream cost of additional support structures, physical handling challenges with oversized modules, increased insurance premiums, and risk associated with cell cracking and wind loads. As it relates to our copper replacement Series 6, once implemented, we anticipate ASP accretion due to increased efficiency, improved temperature coefficient, and a significant reduction in long-term degradation. Importantly, this innovation is driven by efficiency improvements, which result in dilution of our variable and fixed bill of material costs. We expect to capture this ASP accretion as the technology does not significantly impact balance-of-system and development costs or project risks. Finally, it is important to note the opportunities within our technology roadmap.
With cell efficiency entitlement in excess of 25%, coupled with the energy advantages of CdTe, we believe the outlook for our technology platform remains strong. In support of our roadmap over the coming quarters, we anticipate certifying a new world record for CdTe. I'll now turn the call over to Alex, who will discuss our third quarter financial results and fourth quarter guidance. Alex?
Thanks, Mark. Starting on slide eight, I'll cover the income statement highlights for the third quarter. Net sales in Q3 were $928 million, an increase of $285 million compared to the prior quarter. The increase was primarily driven by the sales of certain Japan and India projects, as well as an increase in the volume of Series 6 modules sold to third parties. On a segment basis, as a percentage of total quarterly net sales, our module segment revenue in Q3 was 46% compared to 58% in Q2. Total gross margin was 32% in Q3 compared to 21% in Q2. The system segment gross margin was 33% in Q3 compared to 21% in Q2. This increase was primarily driven by the aforementioned international project sale and the sale of early-stage development assets, including a project entity associated with the General Motors PPA.
This was partially offset by $14 million of performance liquidated damages stemming from underperformance of third-party equipment at several of our legacy EPC projects. Module segment gross margin was 30% in Q3 compared to 22% in Q2. There were several positive and negative factors that impacted this Q3 result. Firstly, we recorded a reduction in our product warranty liability reserve, which was primarily due to lower warranty settlements than previously estimated for our Series 2 technology. This resulted in a $20 million reduction of our warranty liability and a corresponding benefit to cost of sale. Secondly, certain of our legacy module sale agreements are covered by a collection and recycling program where a corresponding expense for the estimated future cost of our obligation was recognized at the time of sale.
In Q3, we recognized a $19 million reduction in our module collection and recycling liability due to changes in the estimated timing of cash flows associated with capital, labor, and maintenance costs. This also resulted in a corresponding benefit to cost of sales. Finally, we incurred an impairment loss of $17 million to certain module manufacturing equipment, including framing and assembly tools no longer compatible with our long-term technology roadmap. This resulted in a corresponding increase to cost of sales. On a net basis, these factors increased module segment gross margin dollars and percent by $21 million and 5% respectively. Our module segment gross margin was impacted by - $6.5 million of Series 4 gross margin, which included $2 million of decommissioning and severance costs. Our Series 4 gross margin reduced overall module segment gross margin by 1.5%.
With these facts in mind, we're pleased with our overall module segment gross margin result and our Q3 Series 6 gross margin relative to our previous expectation of 25%. This was exceeded despite lower than expected Q3 volumes sold and despite incurring $3.5 million of unforeseen COVID-19-driven logistics costs in the quarter. Additionally, and as a reminder, sales freight and warranty are included in cost of sales and reduced module segment gross margin by 6%. SG&A and R&D expenses totaled $73 million in the third quarter, a decrease of approximately $1 million compared to the prior quarter. This decrease was primarily driven by lower severance and project impairment, partially offset by higher legal and incentive compensation expense. Production startup, which is included in operating expenses, totaled $13 million in the third quarter, an increase of $7 million compared to the prior quarter.
This increase was driven by higher startup expense for our second Series 6 factory in Malaysia. Interest income was $2 million in the third quarter compared to $4 million in Q2. This decrease was primarily driven by lower interest rates and investment balance for our marketable securities. We recorded tax expense of $38 million in the third quarter compared to $10 million in Q2. This increase in tax expense is largely attributable to higher pre-tax earnings in Q3. The combination of the aforementioned items led to third quarter earnings per share of $1.45 compared to $0.35 in the second quarter. I'll next turn to slide nine to discuss select balance sheet items and summary cash flow information. Our cash, marketable securities, and restricted cash balance ended the quarter at $1.7 billion, an increase of approximately $29 million compared to the prior quarter.
Total debt at the end of the third quarter was $261 million, a decrease from $465 million at end of Q2 as a result of international project sales. As a reminder, all of our outstanding debt continues to be project related and will come off our balance sheet when the corresponding project is sold. Our net cash position, which includes cash, restricted cash, and marketable securities, less debt, increased by approximately $233 million to $1.4 billion. This increase was driven by proceeds from system sales and module segment operating cash flows, which were partially offset by capital expenditures and loan repayments associated with the Ishikawa project sale. Net working capital in Q3, which includes non-current project assets and excludes cash and marketable securities, decreased by $18 million compared to the prior quarter.
Net cash provided by operating activities was $208 million in the third quarter, compared to $148 million in the prior quarter. As it relates to our Ishikawa project, we repaid the project debt prior to close, which resulted in higher operating cash flows upon transaction close. As it relates to our Miyagi and Anamizu project, the associated project debt was assumed by the buyer, which reduced cash outflows from financing activities and operating cash inflows at transaction close. Finally, capital expenditures were $106 million in the third quarter, which brings our year-to-date total to $327 million as we continue our Series 6 capacity expansion. Turning to slide 10, I'll next provide an updated perspective on 2020 guidance. On our Q2 call, we provided guidance metrics that we believed were largely within our control or within reasonable line of sight at the time.
This included production, operating expense, and capital expenditure guidance for the full year 2020. While significant uncertainties remain regarding the severity and duration of the COVID-19 pandemic and its impact on our operations and financial results, we believe visibility into our financial performance for the fourth quarter has improved on account of the following. At the time of our Q2 earnings call, we cited tax equity market uncertainty for projects set to achieve commercial operation in 2021, which had the potential to impact our module customers and our self-developed Sun Streams 2 project. Provisions for credit losses stabilized somewhat in Q3, significant uncertainty remains for Q4 and 2021. Forecasting tax capacity for the second half of 2021 remains difficult due to the uncertain economic outlook, government response, and trajectory of COVID-19. Some tax equity providers have begun committing to 2021 financings, albeit at conservative volumes.
Whilst this is incrementally positive for the U.S. solar market, we remain strongly supportive of a direct pay legislative solution in lieu of the investment tax credit to alleviate potential disruptions in the tax equity market. Secondly, with the sale of our American Kings project in Q2, the previously mentioned sales of the Japan and India assets in Q3, and with the expected sale of the Sun Streams 2 project in Q4 2020 or Q1 of 2021, we have a viable path to close each project sale contemplated in our original 2020 guidance from February. Finally, at the time of the February guidance call, we anticipated full-year module shipments of 5.8 GW to 6 GW . Shipments through the end of the third quarter total 3.7 GW, which is below our year-to-date expectations, largely due to COVID-19-driven projects, financing, logistics, and customer delays.
This has resulted in a shipments profile incrementally weighted to the second half of the year. A significant volume of modules that we anticipate recognizing as 2020 revenue are currently in transit or will be shipped in the coming weeks. With the improved visibility for system sales for 2020 and year-to-date progress for module shipments, we are reinstating financial guidance for the fourth quarter that considers this range of outcomes for module revenue recognition timing and closing of the Sun Streams 2 sale. Given the uncertainty around any outcome from the evaluation of strategic options for our U.S. project development business and the sale timing of our North American O&M business, our fourth quarter guidance assumes no change to our existing lines of business.
To date, while we've largely managed the impact of COVID-19 on our business and has not had a significant impact on our operations, our guidance assumes it will continue to be able to mitigate any such impacts on our supply chain and operations without the incurrence of material cost. I'll now review our fourth quarter guidance ranges with the implied full year 2020 guidance ranges included on slide 11. Starting with shipments, due to the aforementioned uncertainties, we anticipate volumes of 1.8-2 GW in Q4, which implies 5.5-5.7 GW for the full year. Production in Q4 is expected to be 1.5 GW with implied full year 2020 production of 6 GW , including 0.2 GW of Series 4. Note that our full year production guidance of 6 GW has increased by 0.1 GW relative to the guidance provided during our Q2 earnings call.
Net sales are expected to be between $540 million and $790 million in Q4, which accounts for potential delays in the close of our Sun Streams 2 project sale and module revenue recognition timing. Total gross margin is projected to be 26.5%- 27% in Q4. Note that we anticipate the closing of the Sun Streams 2 project sale will be slightly dilutive to overall gross margin %. Module segment gross margin is projected to be between 27% and 28% in Q4. We anticipate our module segment gross margin will be supported by an increase in volume sold and continued reductions in our cost pool, partially offset by a modest decline in ASP. Included in this gross margin guidance is an anticipated 40 basis point gross margin % drag due to Series 4. Operating expenses are expected to be between $90 million and $95 million in Q4.
This includes production startup expenses related to our second Series 6 factory in Malaysia of $15 million. We anticipate R&D and SG&A costs, excluding startup expenses, of $75 million- $80 million in Q4. Our current implied full year 2020 operating expense guidance of $351 million- $356 million is within the guidance range provided during our second quarter 2020 earnings call. Operating income is estimated to be between $50 million and $120 million in Q4. Turning to non-operating items, we expect interest income, interest expense, and other income to net to - $5 million. We anticipate a tax benefit of $45 million - $60 million in Q4, which includes a discrete tax benefit of $60 million associated with the closing of the statute of limitations on uncertain tax positions.
This results in a Q4 earnings per share guidance range of $1.00 - $1.50 per share, and $3.65 - $4.15 per share implied for the full year 2020. Our 2020 capital expenditure forecast of $450 million - $550 million remains unchanged from the prior quarter. Our year-end 2020 net cash balance is anticipated to be between $1.2 billion and $1.3 billion. We do decrease relative to our Q3 net cash positions primarily due to capital expenditures and project spend related to our Sun Streams 2 project. Turning to slide 12, I'll summarize the key messages from today's call.
Firstly, we had Q3 earnings per share of $1.45, increased our quarter-end net cash position, improved module segment gross margin quarter-over-quarter, and reinstated financial guidance for the fourth quarter. Secondly, we had strong manufacturing performance with each factory averaging over 100% capacity utilization in Q3, and at our Malaysia factory, achieved our midterm cost target of a 40% reduction from our 2016 Series 4 cost. Thirdly, demand for our Series 6 products is robust, and we have continued success adding to our contracted pipeline with net bookings of 1.6 GW since the prior earnings call and 4.1 GW year-to-date. Our contracted backlog remains a pillar of strength with 6.7 GW contracted for expected deliveries in 2021, and 3.6 GW contracted for expected deliveries across 2022 and 2023.
Finally, despite ongoing challenges relating to the COVID-19 pandemic, we remain pleased with all operational and financial performance. With that, we conclude our prepared remarks and open the call for questions. Operator?
As a reminder to ask a question you will need to press star one on your telephone. To withdrawal your question press pound or the hash key. In order to attend everyone to ask a question please keep a limit to yourself for one question. Please stand-by as we compile the Q&A question. Our first question comes from Philip Shen with ROTH Capital Partners. Your line is open.
Hi, everyone. Thanks for the questions. Have a few here, so thank you for your patience. The first one is on CIGS. I think we picked up recently that you guys restarted your CIGS research efforts. Wanted to get a sense for what that might mean, relative to CdTe. Second, I think you guys have talked about getting to 8 GW of capacity by year-end 2021. Can you talk about the conditions that need to exist to consider expanding capacity, and then what the timing and the locations of that might look like? Number three here, with the ever-growing importance of security of supply, especially in the face of the Xinjiang risk and the concentration of capacity in China, can you talk about how the tenor of conversations with customers may have changed over the past month?
Help us understand how much is booked in 2021, 2022, and 2023. Thanks, guys.
All right. I'll let Alex take the last one with the booked in 2021, 2022, and 2023. I'll try to take the other three as he gathers those numbers. Your question around CIGS and where, to the extent that we are looking at that as a technology or really any technology, because, at the end of the day, the core, we're a module technology manufacturing company. That's what we do. We need to continue to find ways to differentiate ourselves on a technology basis and to find ways to be as disruptive as we can with our technology and always increase advantages relative to crystalline silicon and whatever the ultimate competition may be. We've looked at CIGS in the past. We've looked at crystalline silicon and monocrystalline silicon. We look at perovskites. We look at everything.
As it relates to you, if your question is CIGS is somewhat an indication that we don't have confidence in CdTe. That's not anywhere close to the reality. CdTe, in my mind, will always be advantaged relative to CIGS. We believe it has a roadmap to be advantaged relative to crystalline silicon. The reason that we would look at crystalline silicon or CIGS or perovskites or organic PV or anything else that may be out there is how do you complement the two? We've mentioned this before, that we believe over time that multi-junction type of technology will evolve into the marketplace. CdTe, in and of itself, is a very good top cell from a standpoint of having a high bandgap. It captures a significant portion of the overall sun spectrum.
What you need then as a second cell underneath that would be what's complementary to that? Could it be CIGS? Could it be crystalline silicon? Could it be perovskites in some form or fashion? We're always going to look at that. We believe it's the evolution of technology that will happen in the future. We've highlighted in previous calls, too, that we are reinvigorating our advanced technology team. To the extent that we need to look at different materials, different semiconductors that could be complementary to CdTe, we'll continue to do that. I don't look at CIGS or really any of those other technologies as replacing CdTe. The core will be of our semiconductor and the overall device will be CdTe.
The question is there something else that becomes complementary with it, and we need to look at all options that could be out there. As it relates to the expansion in the 8 GW, we have, I think, indicated our nameplate will be about 8 GW at the end of 2021. We are in the last phase of our Series 6 transition, with our second factory in Malaysia that will start up, and current plans will start up in Q1 2021. We are looking beyond that. As we've always said, we've got a balanced business model between growth, liquidity, and profitability. We obviously understand the value of growth and the leverage against fixed costs and flow through contribution margin. We are evaluating options, and we will look at where we can create the next disruptive lowest cost factory.
As we continue down our journey with Series 6, one of the things that we tasked ourselves with. Probably about nine months or a year ago, what's the factory of the future look like? What's the next generation technology in terms of driving the lowest cost for our CdTe platform? We have been doing that and trying to find a way to create that disruptive lowest cost from the future factory, which under that construct, most likely would not be in the U.S. for the reasons that we've mentioned. We believe we have relative representation here with almost 2.5 GW In the U.S. Other markets we would look to, India could be one.
We've talked before about potentially our factory in Germany, starting that up as an easy way to get to market quickly and leverage that factory that we closed down almost a decade ago. Those are things that we're looking at. We also believe that, as we've indicated, I think Alex mentioned in his prepared remarks that, freight is basically 6% of gross margin right now, 6% of the ASP, effectively. Getting closer to market and driving down your freight cost is an advantage that we would continue to look to. Being in a market that can consume 2.5 GW to 5 GW of volume on a recurring basis would be important.
Those are the many different factors that we look to as we think about expansion, but just want to make sure it is clear that is on the agenda of things that we're looking at this point in time. Supply security and any impact with what's happened with China and the latest evolution now being on the potential use of forced labor and implications that that has and potential product being banned in the U.S. It's one of many different dimensions that a lot of our customers have as it relates to relying on China in long term and the uncertainty around that. I don't know if it has increased meaningfully over the last month. It's just one of many different concerns that they've had.
Even with the discussion around bulk power systems and potential implications around that, the modules have been looked at as, is there a risk of that could be impacted? There's a discussion of extension of the 201 tariffs. There's many different things that come into the mix and a lot of our customers, here in the U.S. in particular, want to do business with an American company. They want to de-risk their supply as it relates to ability to deliver against commitments and any uncertainties that could be imposed upon the Chinese supplier and some of the issues that are coming up, as you mentioned with the forced labor. I don't know if it's changed significantly in the last month.
It clearly is top of mind, and it's just one of many different factors that all of our customers think about when they have to rely upon our Chinese counterparty as a supplier. Alex?
Yeah, just to throw the numbers on the end of there. Right now we have 6.7 GW booked for 2021 and 3.6 GW booked across 2022 and beyond.
Our next question comes from Brian Lee with Goldman Sachs. Your line is open.
Hey, guys. Thanks for taking the questions. Congrats on a good quarter. A couple things from my end, I guess. I know there's a lot of moving parts here on the module gross margins this quarter. I'm getting to like a 26.5% clean number for Series 6, which is 150 basis points higher quarter-on-quarter. I guess first question is that the right math? Sort of what happened during the quarter that got you the additional 150 basis points, because I think last quarter you were talking about sort of a flattish sequential trajectory. The second question would be around, I didn't hear it on the call, I might have missed it, but are you still on track to hit that 300 basis point gross margin expansion for Series 6 in Q4 off of this higher run rate, if my math is right?
Maybe just lastly, how should we think about cost reductions in 2021? You said you're running at or above the 10% target for this year. What sort of gross margin expansion potential do you have as you move into 2021, given the cost reduction progress you're making and also the fact that pricing seems to be fairly stable for you guys? Thank you.
Yeah, I'll start on the gross margin and see if Mark wants to comment on the cost reduction beyond that. You're doing basically the right math. With the reported module segment gross margins, about 29.5%, and you've got three big things that moved in there. We had a reduction on warranty liability of about $20 million. That's a good guy. Reduction in the end-of-life recycling of about $19 million, also a good guy. You had an impairment charge of about $17 million that goes against. When you net all of those, you've got about $21 million or about 5 percentage points. You take that down to about 24.5%. We said that within the overall margin, we had about a $6.5 million or 1.5% negative associated with Series 4.
You add that back in, you get to about 26%. Then on top of that, we did have some COVID-19 charges in that number. There's about $3.5 million of COVID-19 charges. You back that out again, you're going to be another 50 to 100 basis points. Somewhere around 26.5 % to 27% is around the Series 6 clean number. Yeah, we're a little ahead of where we expected to be, and part of that is just a function of how well the factories have been running. As I think Mark mentioned in his prepared remarks, we've now managed to get to the cost reduction target that we expected in our high volume manufacturing by the end of the year. In Q3, we're already there in our Malaysia site. We already achieved that last quarter in Vietnam. Running a little ahead there.
In terms of where we think we'll be by Q4, we're going to be at that 27.5%, 28% gross margin on series 6 is the number we're guiding to for Q4. That one tells roughly we're steady from where we were last quarter. Then in terms of cost reduction beyond that, we're not giving guidance out through 2021. Mark, I don't know if you want to comment on any of this.
Yeah, Brian, I guess what I would say just on that is, and it's similar to, I think, what I said in my comments, that the things that continue to drive improvements to the cost per watt are improved throughput, improved watts, and improved yields. What I said in my remarks is that we just opened up the 445 W in. Over the next few quarters, we're going to be at the 455 W and north of that shortly thereafter. That improvement in watts is going to help drive down costs, so that helps. The throughput, as we've already highlighted, is going to continue to increase. That's another significant lever that is going to help drive down the cost per watt.
The other one is, we highlighted, is the glass supply for our Perrysburg factory, which will start to take shape in Q4, and we'll see you realize the full benefit starting in Q1. The bill of material cost in Perrysburg particularly is going to be helpful. We're also working on taking significant cost out of the frame, which will help drive costs further. Then we're actually looking at changing the profile of the module such that it's actually thinner. When you look at the glass plus the frame, we're trying to make the profile of the module thinner, which will also then allow us to add about 10% more modules, a little over 10% more modules per container from a freight standpoint, which will help drive down that cost.
We've got teams that are really multidimensional and looking across all aspects of everything that is associated with the cost of the product and ultimately delivering it to the end customer. As Alex said, we're not guiding at this point in time for 2021. We are tracking to be ahead of where we thought we were going to be on a cost per watt basis as we exit the year, which is a good place to be. We still have a number of other initiatives that will further drive down our cost a watt as we go into 2021.
Our next question comes from Ben Kallo with Baird. Your line is open.
Hey, great quarter, guys. Three or four questions here. Can you talk about your utilization here above nameplate, and how should we expect that going forward? I know, Mark, you just talked about watts and your yield. How much can we go above nameplate capacity? When we think about your target for next year, how much is that considered in that? You don't talk about efficiency anymore. Can you talk to us about the theoretical efficiency of CdTe and where we can go from a 445 W panel and above? The last question, I guess is, the 2.5 GW before you place new capacity somewhere, if it's in Oderson, I get to $400 million of CapEx. Does it take you selling the systems business before you make that go, no-go decision?
Can you talk about just the incentive to onshore new capacity? Thanks.
Ben, what I said in the script was that where we are on trajectory right now for our two specific plants in Ohio is to drive the nameplate capacity up to 25% above what we originally launched at. The objective would be to drive it effectively to about low 30s. Let's say there's another 5 to about 8 percentage points of incremental utilization that we would like to get out of all the factories. First we had to get everything up to 25%, right? We're not necessarily there yet. You can see by what we're running, call it 120s, I think is the best that we highlighted right now.
We've got a path to get everything up to around 25% near term with a goal of getting it up to low 30s, call it around 30%-33%, which would be another 5 to 8 percentage points above what our near-term objective is. A little bit of work still to be done there, but been very impressed with the team's capabilities of making that happen. On the efficiency side, as you indicated, where are we going with the technology? What I did say in one of my very last comments was the entitlement at the cell level for CdTe is 25%+. That would translate into something close to, call it a 23% or so efficient module based on normal conversion between cell to module. We're in the process right now of validating, and hopefully we'll have over the next few quarters, a new record for efficiency.
I don't know if it'll be a cell or a module. Still looking at what we end up doing on that. There's some items on our roadmap that would really be beneficial to help validate that through demonstrating that either with a cell or a module, which by default would then drive to a record efficiency. There's still more room to go yet on the core technology and driving to higher level efficiencies.
The other one that I mentioned that again, it's complementary, and it's a combination, this multi-junction technology that we talked about, which would take efficiencies even above and beyond that profile, as I mentioned, because now you have two cells that are harvesting the overall photons and converting them into electrons, which would say that whatever you can do at the top cell level capability would be further enhanced with the bottom cell. The technology on efficiency can go even beyond those numbers that I represented around theoretically. There's still a ways to go yet on the technology, and we're working very hard to make sure we can realize those benefits over the next several years.
On a CapEx basis, as I mentioned, I alluded to this, what is the factory of the future and what is the most disruptive, lowest-cost factory that we can create, starting with a blank sheet of paper. That'll also drive down the CapEx investment, relative to what we initially communicated with Series 6. That can drive it down probably close to 15%-20% on a CapEx basis. We have a roadmap to drive more CapEx out to be more efficient, so therefore the cost to deploy a new factory on a CapEx per watt basis will be lower than when we originally launched that, which is an objective that we've had. As it is tied to the realization of the systems/energy services and proceeds associated, they're two independent things. There's not a constraint.
You can see by how our balance sheet is right now, that ability to continue to invest and the capacity and the roadmap that we have in front of us, there's not any linkage per se to any proceeds we receive, either from the systems business or the energy services business.
Our final question comes from Michael Weinstein from Credit Suisse. Your line is open.
Hi, guys. Can you talk about what would drive capacity adds in any particular regions or existing locations? In the past few years, this has been driven by a move from the Series 4 to Series 6. Going forward, is it demand growth or something else that drives it?
Look, whenever we grow, we want to make sure that it's tied to a market that would be attractive to us. Again, being closer to the market or at least driving efficient shipment logistics routes to access a market is important. The reality is freight cost is north of 10% of the overall cost of the product. If you can get that number down, say, cut it in 1/2, you get 5% savings just by getting closer to the customer and reducing your overall freight cost to deliver the technology. There's many other factors that we use when we screen a site. Energy is a meaningful component of our overall cost, so we have to have competitive cost of energy. Labor rates have to be competitive from that standpoint as well.
While the processes are largely automated, it's still a major component of the overall cost structure that has to be thought through whenever we make decisions of where we're going to manufacture. The reality is that we will grow. Most likely, it's probably going to be outside of the U.S. as we currently envision it. It could change. We have ability just to expand within the footprint and drive more throughput of what we already have here in the U.S. as we think about where our next factory would be.
As we currently envision it would probably be in an international market somewhere close to customers, close to where there's a strong recurring annual demand requirement and one that our technology's well-positioned and competitively advantaged, whether it's hot humid climates, whether it's the advantages of our CO2 footprint, environmental aspects around our technology, which more and more markets are starting to value. Those are the types of things that we take into consideration as we would think about any expansion.
We have completed the allotted time for questions. This concludes today's conference call. You may now disconnect.