2026.10.09 | 9K | Power | FTAI Aeroderivative Progress Channel Check

音频信息

音频标题: 261009 - Merit ftai近况及航改燃进度调研

相关标的：undefined

议题：

完整音频链接：http://mw-meeting.oss-cn-shanghai.aliyuncs.com/audio/1791516004682-261009%2520-%2520Merit%2520ftai%25E8%25BF%2591%25E5%2586%25B5%25E5%258F%258A%25E8%2588%25AA%25E6%2594%25B9%25E7%2587%2583%25E8%25BF%259B%25E5%25BA%25A6%25E8%25B0%2583%25E7%25A0%2594_20261009_111856.mp3

音频纪要总结

专家访谈纪要：FTI Mod 1航改燃机项目进展、量产能力与供应链分析

1. 项目参与程度、当前阶段与首次交付定义

a) 专家在Mod 1项目中的参与程度如何？

• 专家直接参与了Mod 1项目的多个核心环节，包括：

1. 发动机配置，确保设备在配置层面正常运行
2. 部分设计工作
3. 原型机测试及测试问题排查
4. 硬件选择和工程决策

• 除市场推广和最终包装环节参与较少外，专家基本接触了项目的各个技术环节。
• 专家来自蒙特利尔工厂，并持续与发动机及零部件团队合作，对制造、返厂维护和测试进展较为熟悉。

b) Mod 1目前处于什么阶段？

• 当前仍处于测试、运行包线定义和控制参数微调阶段。
• 首台Mod 1正在蒙特利尔工厂制造，相关测试仍在佛罗里达州Medley/迈阿密测试设施进行。
• 项目已经完成主要设计验证，不再处于可能全面推翻设计的早期阶段。
• 当前工作的重点包括：

1. 确定发动机可行运行范围
2. 微调瞬态运行参数和控制逻辑
3. 开展耐久性和长时间连续运行验证
4. 为设计最终定型及扩大生产做准备

c) 管理层所称“2026年第四季度首次交付”具体指什么？

• “首次交付”主要指发动机具备投入运行的能力。
• 这一节点不一定意味着整套设备已经完成最终拖车式封装，并在客户现场完成全部验收。
• 其实际定义介于以下两种状态之间：

1. 产品已达到可以运行、基本准备就绪的状态
2. 产品已经完成客户验收并交付至客户现场

• 因此，管理层的首次交付口径更接近“发动机或核心设备具备商业运行条件”，而非严格意义上的整套系统最终现场验收。

d) 2026年第四季度首次交付目标是否仍然可行？

• 专家认为目前没有已知的重大重新设计事项会影响2026年第四季度首次交付。
• 第一次迭代版本的主体设计已经得到验证，目前主要是在确定并微调运行包线。
• 项目已经远离“测试结果可能全部失效、需要推倒重来”的阶段。
• 早期VBV控制问题已在数月前完成主要攻关，目前不会对现有原型机和首次交付时间表构成明显影响。
• 综合判断，2026年第四季度首次交付目标仍然具备可实现性。

2. CFM56改造成Mod 1的技术路径

a) CFM56改造成Mod 1需要进行哪些改动？

• Mod 1约90%的CFM56发动机部件可以重复使用。
• 除燃烧部分和少量细节调整外，核心机周边的大部分结构基本保持不变。
• 第1代Mod 1中，低压涡轮（LPT）基本不变，主要改造集中在发动机前端和燃烧部分。
• 核心改动包括：

1. 重新设计风扇盘
2. 重新设计并制造风扇叶片
3. 在增压级前增加进口导向叶片，以调节气流
4. 重新设计增压器相关叶片和外侧罩环
5. 取消或弱化与涵道气流、航空推力相关的功能
6. 调整燃油喷嘴、燃油歧管和燃烧室衬套
7. 增加发电应用所需的仪器设备
8. 拆除原本服务于短舱和飞机附件的部分线束
9. 重新开发与进气变化相匹配的控制逻辑

b) 哪些部件是重新设计和验证的重点？

• 发动机前端是重新设计和重新制造工作最集中的区域，主要包括：

1. 风扇盘
2. 风扇叶片
3. 增压器进口导向叶片
4. 增压器一级或二级叶片

• 风扇盘在航空应用中属于寿命限制件，通常需要严格测试、分析和寿命计算。
• 在发电应用中，项目更关注蠕变问题，而不是航空发动机频繁起降所导致的循环寿命，因此可以采用不同设计规范，并保留更高工程裕度。
• 发动机热端的主要改造对象包括：

1. 燃油喷嘴
2. 燃油歧管
3. 燃烧室衬套

• 整体来看，增压器和燃烧系统是Mod 1最主要的两大改造区域。

c) 为什么90%的原有部件可以继续使用？

• CFM56核心机及主要结构已经经过长期航空应用验证，具备成熟的可靠性基础。
• 对于保持不变的部分，维护和维修可以基本沿用OEM指导规范。
• Mod 1不需要满足完整航空适航要求，因此部分原本服务于航空推力和适航维修性的标准可以重新评估。
• 例如，动力发动机不依赖涵道气流产生航空推力，因此部分导向叶片不必保持与航空应用相同的可维修性标准。

d) Mod 1是否全部采用全新零部件制造？

• 不会。Mod 1主要采用以下组合方式：

1. 对现有CFM56发动机进行改装
2. 使用可修复和再制造零部件
3. 对关键改造件采用新制造零部件
4. 从OEM或PMA供应商采购缺口部件

• 项目并不是以全新零部件重新制造一台发动机，而是通过改装、再制造和供应链采购相结合的方式完成组装。
• 这一模式是Mod 1成本、交付速度和产能可行性的基础。

3. 测试进展、运行包线与关键工程问题

a) 当前测试硬件与最终商业产品有多接近？

• 当前测试设备已经非常接近Mod 1第一次商业化迭代版本。
• 现阶段不存在可能导致项目全面重做的重大设计事项。
• 主要工作已从“验证设计是否可行”转向：

1. 确定运行极限
2. 微调运行包线
3. 优化控制参数
4. 验证长时间连续运行能力
5. 提升效率并降低排放

• 后续第2次和第3次迭代主要侧重效率和排放优化，而不是证明基础架构能否运行。

b) 目前有多少台原型机，累计进行了多少测试？

• 具体原型机数量未披露，但项目拥有多台原型机，并通过轮换方式开展测试。
• 原型机停机期间会进行拆解、维护和检查，以验证其表现是否符合仿真结果和设计预期。
• 专家深度参与过其中一台原型机的测试。
• 在专家参与期间，单次最长连续运行时间约为4至5小时，但并非全程处于峰值负荷。
• 测试需要覆盖不同负荷、加减速状态和运行包线，因此不会一直维持最大功率。

c) 下一阶段最重要的测试里程碑是什么？

• 下一项关键里程碑是连续运行一周的耐久性测试。
• 一周连续运行后，团队将拆解和分析发动机，重点检查此前可能未充分暴露的额外蠕变问题。
• 主要观察部件包括：

1. 低压涡轮第2级叶片
2. 高压涡轮叶片

• 项目团队希望通过试验和计算确认：

1. 发动机能够在预期工况下长时间运行
2. 设备可以连续运行更长时间，例如一个月，而无须计划外停机
3. 关键热端部件能够达到预期全生命周期要求

• 访谈中并未确认已经完成一周全天候连续运行；连续一周运行仍是近期测试目标。

d) 测试中最主要的故障排查问题是什么？

• 测试期间最主要的问题集中在VBV可变放气阀系统及其控制逻辑。
• VBV阀门硬件继续采用OEM零部件，变化主要来自控制系统和逻辑参数。
• 由于Mod 1采用不同的增压级和进口导向叶片设计，进气流发生变化，进而影响：

1. N1低压转子响应
2. N2核心机响应
3. 燃油输入量
4. 发动机启动、加速和减速过程

• 如果燃油输入过多，而VBV控制未正确匹配，核心机可能失速；如果控制系统识别到异常，也可能要求发动机停机。
• 该问题本质上是复杂的PID控制和运行逻辑匹配问题，需要通过反复测试界定不同转速、加速度和燃油输入之间的关系。

e) VBV问题是否会影响首次交付时间表？

• 不会构成当前时间表的主要风险。
• 相关集中测试持续约2.5周，主要工作已在数月前完成。
• 团队已经明确不可进入的运行区域和主要运行极限。
• 目前剩余工作主要是微调运行包线，使发动机在效率基准和运行指引内工作，并控制燃油喷射量、超调、堵塞和供给不足等问题。
• 因此，VBV问题已经从关键技术障碍转变为参数优化工作。

f) 当前最主要的耐久性风险是什么？

• 当前最主要的潜在耐久性风险是蠕变和高周疲劳。
• 低周疲劳主要由热循环驱动，可以根据循环次数和运行规则进行管理。
• 团队重点关注低压涡轮第2级叶片，因为：

1. 该部件长期处于高温环境
2. 原设计并非单晶材料
3. 虽有防蠕变和扭转变形涂层，但仍需进一步确认全生命周期表现

• 当前采取的措施包括：

1. 将低压涡轮第2级叶片更换为新品，建立测试基准
2. 开发单晶叶片替代方案
3. 检查高压涡轮叶片是否存在蠕变裂纹扩展

• 目前尚未发现随机开裂或失效迹象，但这两类叶片仍是后续测试重点。

4. 性能、效率与排放情况

a) Mod 1当前实际输出功率是多少？

• 管理层目标为至少25MW。
• 在燃烧室尚未进一步优化、VBV调度也未充分优化的情况下，首台原型机已经实现约21MW的总输出功率。
• 专家预计实际商业产品输出将接近25MW，至少高于20MW。
• 当前测试为完整动力链测试，覆盖：

1. 发动机
2. PTO轴
3. 齿轮箱
4. 发电机

• 因此，21MW并非单纯发动机轴功率，而是完整系统测试中的总输出结果。

b) 25MW目标是否合理？

• 专家认为25MW是合理且可实现的目标。
• 当前测试地点位于迈阿密，高温和高湿度属于较不利运行条件。
• 在低温、干燥环境下，发动机具备更高的EGT裕度，可以在相同排气温度约束下以更高转速或更高负荷运行。
• 因此，在优于迈阿密的环境条件下，设备输出可能进一步提高。
• 最终产品竞争力不只取决于能否达到25MW，还取决于：

1. 达到25MW所需的燃料量
2. 热耗率和系统效率
3. 排放水平
4. 连续运行可靠性

c) 当前效率和热耗率表现如何？

• 当前发动机热耗率和效率总体符合预期。
• PTO轴、齿轮箱和发电机等部件会带来一定效率损失，但整体仍在项目目标范围内。
• 访谈未披露具体热耗率或效率数值。
• 后续迭代将继续提升整体效率并降低单位发电燃料消耗。

d) 当前排放控制方案是否已经定型？

• 排放控制的核心在燃烧系统。
• 目前考虑的方案包括：

1. 双喷射
2. 水冷或喷水方案
3. 燃烧室进一步迭代
4. 外部SCR系统

• 当前燃烧室采用第一次迭代设计，其排放水平接近CFM技术插入型发动机中效率较高的版本。
• 第一次迭代的主要目标是证明航改燃机可以稳定运行并达到25MW。
• 第2次和第3次迭代将进一步优化效率和排放。
• 燃烧室排放约束是最可能造成后续迭代延误的因素，但预计不会导致整体生产计划推迟6个月。

e) 噪声与认证风险如何？

• 主要监管和认证关注点包括：

1. 排放
2. 噪声
3. 比油耗或SFC

• 原型机过去数月一直在迈阿密进行测试，基础原型状态已经基本满足当地城市噪声标准。
• 专家认为噪声达标风险较低，在美国西部工业区域等应用场景中应更容易满足要求。
• 对具体排放法规和认证要求，专家参与有限，无法给出确定判断。
• 若CFM56技术插入型发动机的排放可以配合SCR满足要求，Mod 1预计能够达到相同或更低排放水平。

5. 2027年100台产能目标与工厂资源分配

a) 2027年生产100台Mod 1是否可行？

• 专家认为该目标具有挑战性，但总体可行。
• 蒙特利尔工厂每季度大约处理130个主要气动模块。
• 扣除约30至40个单独销售模块后，其余模块大致相当于每季度约30台发动机，即每年约120台。
• 这一估算尚未计入：

1. 迈阿密工厂每季度约15台的潜在处理能力
2. 罗马设施每季度约10台的潜在处理能力

• 虽然维修与制造、改装的具体流程不同，但现有处理规模表明公司具备实现年产100台的基础能力。

b) 实现100台目标的关键条件是什么？

• 主要条件包括：

1. 建立严格、标准化且高效的发动机改装流程
2. 建立稳定的拆解线和再制造体系
3. 保证可修复零部件的供应
4. 补足OEM、PMA和新制造零部件的缺口
5. 合理配置人力和设备
6. 协调航空业务和发电业务之间的资源

• 公司正在采购部分发动机用于拆解，并利用不可用材料或剩余寿命较低的材料重建发动机。
• 真正的挑战不是单纯装配能力，而是建立高效的物料吞吐和转换体系。

c) 传统航空业务与发电业务将如何分配产能？

• 蒙特利尔工厂目前拥有两条主要装配线，其中较新的装配线主要服务于发电业务。
• 发电业务在爬坡阶段仍会占用部分航空业务的人员、供应链和设施资源。
• 专家预计未来6个月内，蒙特利尔工厂的业务分配可能逐步接近50/50。
• 已签订的发动机交付合同会获得较高优先级。
• 发电业务利润率更高，公司在确定资源优先级时预计会向发电业务重新分配资源。

d) 发电业务是否会挤压航空业务产能？

• 航空业务产出量可能不会维持原有水平，但进入车间的发动机投入量可能持平甚至增加。
• 部分航空发动机因核心机需要较大修复，可能成为账面沉淀资产，但其前端、风扇框架和低压涡轮仍具备较高利用价值。
• 发电业务能够承受更高的核心机修复成本，因此可以将这类发动机拆解、修复并转用于Mod 1。
• 未来车间中可能出现更多专门用于拆解和发电业务转换的发动机。

6. 零部件供应链与叶片瓶颈

a) 新设计零部件的交付周期如何？

• 新设计的风扇盘、风扇叶片和增压级叶片已开始到货。
• 风扇盘约在6个月前完成设计，现场已持有多件产品。
• 在供应商已经掌握全部图纸和规格的情况下，订单交付周期约为2至3周。
• 单件制造及交付周期约为2周，相关供应商产能约为每周2套。
• 供应商已承诺在11月中旬达到既定产出水平。

b) 为什么此前市场认为交付周期需要2至3个月？

• 交付周期取决于项目所处阶段。
• 在设计尚未定型、供应商尚未配置产能时，迭代件交付可能需要更长时间。
• 一旦满足以下条件，交付周期可以明显缩短：

1. 设计完成最终定型
2. 测试包线得到验证
3. 供应商预期会持续获得重复订单
4. 供应商提前配置人力和生产能力

• 因此，2至3周主要是量产准备完成后的重复订单交付周期，而不是所有开发阶段的统一周期。

c) 风扇叶片和增压级叶片是否会制约2027年爬坡？

• 专家判断，风扇叶片和增压级叶片目前不构成主要瓶颈。
• 两类零部件都已取得实物，供应商计划在11月中旬达到既定产量。
• 风扇叶片将采用新制造产品，不会直接使用旧叶片。
• 当前需要持续关注的部件是低压涡轮第2级叶片，但其市场需求较低，可以直接从OEM采购新品，短期采购难度不大。
• 长期替代方案是采用单晶低压涡轮第2级叶片，但暂未确定完成时间。

d) FTI如何规避航空行业常见的叶片供应瓶颈？

• 航空公司通常受既定维修规范限制，部分航空公司拒绝使用任何PMA叶片，因此只能进入OEM等待名单。
• FTI在发电业务中具有更高灵活性，可以同时使用：

1. OEM零部件
2. PMA零部件
3. 内部维修件
4. 再制造件

• 公司与OEM签有MCPS协议，可以获得较优价格，并在部分叶片套件释放时获得一定优先权。
• PMA合作方包括：

1. Pacific Aero
2. Chromalloy
3. Heico

• 公司已收购Pacific Aero，增强了内部叶片生产和维修能力。
• 因此，翼型件供应基本得到落实，风扇叶片供应也预计在11月中旬到位。

e) 哪些零部件仍然存在较强供应限制？

• 真正供应紧张的部件主要是：

1. 高压涡轮叶片
2. 寿命限制件（LLP）

• 这类部件的制造流程严格，涉及专有工艺、材料控制和寿命验证。
• CFM供应仍有较长等待名单，部分套件每3至4周只能满足约30%的需求。
• 第三方很难迅速建立寿命限制件制造能力，因为制造工艺变化会直接影响零部件寿命。
• Chromalloy获得高压涡轮叶片FAA批准约耗时4至5年，表明新进入者难以在2至3年内快速形成成熟供应能力。

7. 单机成本、PMA降本与经济性

a) 一台Mod 1的发动机及系统成本大约是多少？

• 专家仅能估算发动机和主要设备成本，无法确认整套封装系统的准确成本。
• 初步成本拆分如下：

1. 适合发电应用的二手CFM56发动机采购成本：约400万美元
2. 核心机维修和修复成本：约300万美元
3. 机组附件和QEC：约100万美元
4. PTO轴、发电机、齿轮箱及封装等：暂估约200万美元

• 按上述估算，单套Mod 1总成本约为1,000万美元。
• 相比之下，状态较好、航空应用价值较高的CFM56发动机价格约为1,000万至1,200万美元。

b) 为什么发电应用可以采购更便宜的发动机？

• 航空应用更重视剩余循环次数和低周疲劳寿命。
• 发电应用以长期连续运行为主，更关注蠕变而不是频繁起降导致的循环疲劳。
• 因此，Mod 1可以采购循环次数较多、航空价值较低但适合连续发电的发动机。
• 公司可以将节省的采购成本投入核心机修复和关键零部件更换，同时保持更高利润率。

c) 使用PMA零部件可以降低多少成本？

• 不同零部件的降本幅度差异较大。
• 以VBV系统滚珠丝杠作动器为例：

1. OEM认证翻修件约1万美元
2. PMA翻修件约2,000至3,000美元

• 对高压涡轮叶片等昂贵部件，PMA与OEM的价格差距虽没有上述案例悬殊，但仍然显著。
• 若整套HPT叶片价格约为200万美元，采用PMA方案可能节省接近100万美元。
• 对部分高价值部件，PMA价格相较OEM可获得约40%至50%的折扣。
• 寿命限制件不会轻易偏离原有标准，相关部件仍将保持更严格的OEM或合规要求。

8. 客户验证、商业接受度与市场竞争

a) 客户接受Mod 1前需要看到哪些验证结果？

• 保守客户可能同时要求：

1. 完整工厂测试结果
2. 更长时间的现场连续运行数据

• 客户重点关注：

1. 设备能否连续运行至少一周
2. 是否可通过计算和数据证明连续运行一个月而无须停机
3. 发动机监测是否方便
4. 持续运行和维护是否可控
5. 产品是否容易与现场系统集成

• 当前基础工厂测试和气动部分验证已取得实质进展，长时间动力运行测试将是下一阶段重点。

b) 现有客户是否参与测试？

• 已有部分对产品感兴趣的客户参与测试，其中包括已确认订购并为首批设备投入资金的客户。
• 让客户现场了解测试程序，有助于展示：

1. 公司对关键参数的控制能力
2. 测试流程的严谨性
3. 在规定时间内交付最小可行产品的能力
4. 发生问题后的排查和修复机制

• 对于新产品而言，仅凭测试报告难以完全证明长期可靠性，通常需要1至2年的实际运行数据才能充分验证其市场适配性。

c) Mod 1对客户的核心吸引力是什么？

• 当前成熟大型燃机的交付周期普遍较长，采购西门子等成熟产品可能需要至少2年。
• 对迫切需要新增电力吞吐能力的客户而言，主要选择是：

1. 等待2年获得成熟、可靠的传统产品
2. 在未来数月内获得Mod 1并提前满足能源需求

• 即使Mod 1初期在效率、支持服务或吞吐能力方面未完全达到客户预期，客户仍可在等待传统产品期间使用该设备。
• 若Mod 1在价格、效率和模块化方面表现超预期，客户则可能继续追加订单。

d) Mod 1相较潜在新进入者有何优势？

• FTI已经具备：

1. 可供改装的CFM56产品基础
2. 对发动机限制条件的深入了解
3. 3至4个测试设施
4. 大量历史测试数据
5. 发动机维修、拆解和再制造能力

• 从零开发航改燃机的最大难点不一定是硬件，而是控制系统和运行包线。
• 如果不了解发动机在不同转速、负荷和瞬态状态下的运行特性，很难快速建立稳定产品。
• 因此，即使埃隆·马斯克等潜在竞争者计划自研航改燃机，也很难在未来1.5至2年内迅速建立完整零部件、控制系统和量产体系。

e) 新进入者能否通过自制叶片快速解决供应问题？

• 普通翼型件市场已有多家PMA供应商，新进入者即使形成能力，也可能只是新增一家PMA供应商。
• 真正短缺的是高压涡轮叶片和寿命限制件。
• 非型号合格证持有人需要向FAA证明产品符合持续适航要求，认证过程复杂。
• 以Chromalloy为例，高压涡轮叶片获得FAA批准约耗时4至5年。
• 因此，新进入者很难在2至3年内建立可与OEM相当的高价值关键件制造能力。

9. FTI与Jera的职责分工

a) FTI和Jera分别负责哪些工作？

• FTI负责：

1. 在蒙特利尔生产和组装发动机
2. 在迈阿密将发动机与PTO、发电机和齿轮箱结合测试
3. 最终系统测试
4. 测试问题排查与修复

• Jera负责：

1. 将设备集成并封装为移动式拖车发动机系统
2. 完成外部壳体和拖车式结构

• 整体流程为：

1. 蒙特利尔制造发动机
2. 迈阿密进行动力链测试
3. 运送至Jera进行最终封装
4. 再运回FTI完成最终系统测试

b) Jera是否能够支持2027年100台产能目标？

• 专家对Jera封装环节的参与有限，但目前了解的进展符合预期。
• 当前主要工作是准备驱动装配线所需的材料。
• 封装工作尚未正式大规模开始，因此Jera仍有时间调整流程和配置能力。
• Jera在封装流程调整方面效率较高，现有工厂能力基本符合FTI要求。
• 当前没有证据显示Jera会成为2027年100台目标的主要瓶颈。

10. 项目延期风险与替代能源竞争

a) 如果首次交付或产能爬坡推迟6个月，最可能的原因是什么？

• 造成6个月严重延期的最可能原因，是此前未预见的重大蠕变失效。
• 这类问题可能要求：

1. 重新设计关键热端部件
2. 重新制造零部件
3. 重新开展长时间耐久性测试
4. 重新定义运行包线

• 由于项目已经进入最终定型阶段，只有重大重新设计才可能导致6个月级别延期。
• 专家认为发生这种情况的概率较低。

b) 更现实的延期因素是什么？

• 更可能造成局部延期的是燃烧室排放约束。
• 其影响更可能表现为：

1. 第2次或第3次迭代推迟数月
2. 效率优化进度放缓
3. 排放控制方案需要更多测试

• 第1次迭代仍可用于证明航改燃机能够稳定运行并达到25MW，因此燃烧室优化不太可能使整个生产计划推迟6个月。

c) Crusoe取消Boom订单对Mod 1有何启示？

• 客户对交付时间高度敏感。
• 如果供应商无法按照客户预期时间提供产品，客户会迅速寻找替代方案。
• Mod 1的主要优势之一，是相较其他航改燃机方案具备更快的交付速度。
• 潜在风险除交付速度外，还包括：

1. 实际效率不及预期
2. 现场支持不足
3. 长期可靠性验证不足
4. 出现成本更低的新型能源方案

d) 氢能等替代能源是否构成重大风险？

• 当前氢能的能量密度和产业成熟度尚未达到与航改燃机同等水平，短期内不构成直接重大威胁。
• 长期而言，大型能源用户可能采用混合能源方案，包括：

1. 航改燃机
2. 太阳能
3. 电池储能
4. 氢能
5. 煤炭或其他传统能源

• 客户并不执着于某一种技术路线，核心需求是快速、可靠、低成本地获得电力。
• Mod 1长期竞争力将取决于交付速度、可靠性、效率和总体能源成本。

11. 传统航空业务中的PMA零部件趋势

a) 航空客户对PMA零部件的接受度如何变化？

• CFM56正逐步进入生命周期后期，客户对低成本维修方案的需求持续上升。
• 随着OEM减少相关零部件生产或供应，航空公司不得不寻找PMA替代方案。
• 过去一年，航空公司对PMA零部件的态度明显改善。
• 部分过去从不考虑PMA的航空公司，已经开始采购配置全套PMA零部件的发动机。
• 已有航空公司采购包含以下PMA部件的发动机：

1. HPT NGV
2. HPT叶片
3. LPT第1级NGV
4. HPT护环
5. 压气机叶片等其他部件

b) 航空公司如何逐步验证PMA零部件？

• 航空公司通常会先将采用PMA零部件的发动机作为机队先导样本。
• 其核心是观察该发动机能否达到与其他发动机相同的可靠性。
• 若先导发动机表现稳定，航空公司对PMA高价值零部件的接受度可能进一步提升。
• 当前部分航空公司仍倾向先采用重要性相对较低的PMA件，例如起动机，再逐步扩展至核心热端部件。

c) 为什么PMA在航改燃机领域更具吸引力？

• 航空公司对PMA的主要顾虑来自飞行安全和既定维修规范。
• 工业发电和航改燃机不用于飞行，因此对适航和飞行安全的限制明显较少。
• 在航改燃机领域，PMA可以带来：

1. 更低采购成本
2. 更短交付周期
3. 更灵活的供应商选择
4. 更高维修和改装利润率

• 因此，PMA零部件在Mod 1项目中的应用吸引力明显高于传统航空场景。

结构化总结表格

新版prompt提炼之后的转译

提问者：您在 Mod 1 项目中的参与程度如何？具体直接参与了哪些工程或测试工作？

回　答　者：我直接参与了多个环节，包括发动机配置工作，以确保其在配置层面正常运行；也参与了设计和目前仍在进行的测试。总体而言，除市场推广和包装环节参与较少外，我接触了项目的各个方面。

提问者：Mod 1 目前处于什么阶段？是在进行发动机测试、整套发电机组测试，还是已在某个现场运行？

回　答　者：从技术层面看，目前仍处于测试的调节阶段。我能透露的信息仅限于此。FTI 公开发布的大部分信息目前仍然有效。首台 Mod 1 目前正在蒙特利尔工厂制造，测试工作仍在佛罗里达州的 Medley 现场进行。

提问者：管理层曾表示将在 Q4 26 实现首次交付。实际所指的“首次交付”是什么？是设备在客户现场运行、完成发运或调试，还是获得客户验收？

回　答　者：“首次交付”是指发动机能够投入运行，但据我了解，这并不一定意味着已完成整套包装并交付。其状态介于客户验收与设备完全准备就绪、投入运行并交付至客户现场之间。

提问者：在首台机组准备就绪之前，还需要完成哪些工作？哪些关键步骤决定时间表？哪些因素可能延缓交付或产能爬坡？

回　答　者：首台机组准备就绪前仍需完成大量工作。由于这是一款航空衍生型发动机，因此发动机中所有不同于原型的部分都需要进行概念验证。具体到 Mod 1，增压器以及发动机 N1 部分的相关设计有所不同，燃烧部分也存在较大差异，因此这些部分仍需完成验证测试。

此外，还需要完成发动机运行包线的定义，并据此建立相应的控制系统，确保发动机能够在预定运行包线内运行。对于首批机组而言，这些工作是决定时间表的主要因素，首台机组的关键则是运行包线。

从长期可持续性来看，供应链以及严格且灵活的物料吞吐流程将决定发动机的交付进度，因为需要通过该流程将航空发动机转换为航空衍生型发动机。

提问者：请介绍一下将CFM56改造成Mod 1时需要进行哪些改动，哪些部件会被重复使用、改装或重新设计？

回　答　者：CFM56的主要部分都会被重复使用。除燃烧部分和少量细节调整外，核心机周围的部件基本保持不变。将航空发动机改造成Mod 1，实质上可以重复使用CFM56发动机约90%的部件。

在第1代Mod 1中，LPT基本不变，但发动机前端需要重新设计。据我了解，这一设计已经在投资者电话会议上公布。具体而言，风扇叶片有所不同；增压级前方会增加进口导向叶片，用于调节气流；与涵道气流相关的功能基本会被取消；还需要增加额外的仪器设备，同时可以拆除通常用于短舱和飞机附件的其他线束。

总体而言，3个重点改动涉及增压级、风扇叶片、燃烧部分，以及与控制系统相关的所有部件。

提问者：在Mod 1使用的这些部件中，有多少可以采用经过验证的硬件，又有多少需要重新开发和测试？

回　答　者：您的第1个问题是有多少部件可以重复使用，对吗？

提问者：是的。设计Mod 1时需要进行改动和重新设计。我想了解，有多少材料或部件可以继续使用可靠的供应链，有多少需要重新开发，甚至引入新的供应链，因为这可能比较耗时。具体而言，在这些改装或新增部件中，有多少可以重复使用经过验证的硬件或部件，又有多少需要重新设计和测试？

回　答　者：如前所述，CFM56发动机约90%的部件可以重复使用。从维护角度看，基本可以直接沿用OEM的指导规范。对于不需要满足适航要求的部分，也可以从工程角度适当调整。

例如，涵道气流会流经OGV。对于动力发动机而言，这些OGV不需要具备可维修性，因为不会利用涵道气流产生推力。因此，对于这类小型部件，可以重新审视其可维修性标准。

总体而言，发动机约90%的部分保持不变。对于新部件或重新设计的部分，主要部分需要投入最多的设计和验证工作。

回　答　者：风扇盘经过了重新设计。通常，这是一个重要环节，因为从技术上讲，对于航空产品而言，风扇盘属于寿命限制件。因此，需要进行严格的测试、分析和计算，以确保达到规定阈值。在本项目中，我们主要关注蠕变问题，因此设计规范有所不同。不过，我们可以在设计上采用更高的裕度，这也提供了一定的灵活性。

此外，风扇叶片也需要重新设计并重新制造。增压器的进口导向叶片同样经过了重新设计，其外侧还有一个罩环。增压器的1级叶片也需要重新设计，严格来说是2级叶片。因此，发动机前端是重新设计和重新制造工作的重点，需要考虑的相关流程最多。

发动机后端或热端目前涉及燃油喷嘴。后续还将涉及燃油喷嘴、燃油歧管，并可能涉及燃烧室衬套，因为Mod 1将有多个迭代版本，或者更准确地说，会有多种配置。因此，增压器和燃烧部分是2个主要改造区域。

提问者：对于这些重新设计的零部件，是否存在任何限制？这些零部件的交付周期如何？

回　答　者：我们主要与具备相关零部件设计能力的第三方制造商合作。以风扇盘为例，我们现场已经有一些产品，并且已经制造了多件。据我了解，这些零部件大约在6个月前完成设计。每件产品的制造及交付周期约为2周。

下达订单后，只要第三方供应商掌握全部图纸和规格，我们基本可以在2到3周内拿到产品。其产能约为每周2个LP。

提问者：您提到交付周期为2到3周，但在明年进行产能爬坡时，交付周期是否仍然可以保持在2到3周？此前我认为，以风扇叶片为例，交付周期可能需要2到3个月，但您现在提到的是2到3周。因此，从您的表述来看，供应链限制似乎没有我原先预想的那么多。对此应如何理解？

回　答　者：这取决于咨询的对象，以及所询问的项目阶段。我目前具体谈的是Mod 1的第1次迭代，因为Mod 1存在多种配置。就风扇盘而言，正如我刚才提到的，我们现场已经有一些产品。我来自蒙特利尔工厂，需要特别说明的是，这些零部件已经运抵现场。

回　答　者：我每天都在与他们合作。上周，有1台发动机从迈阿密测试设施返回，该发动机已在那里完成测试。新设计的风扇盘已经运抵现场，相关的增压级叶片或风扇叶片也已到位。

由于蒙特利尔目前仍在进行气动构型测试，对迭代件的需求相对较少，因此我们准备了2套。测试完成后，目标是在11月拿到这些部件，这可能与您听到的时间一致。

一旦设计最终定型，并且能够证明测试包线正确，基本上就可以开始扩大生产。通常情况下，如果供应商已经预期会收到重复订单，便可在2、3周内交付部件。举例而言，这个数字仅供参考：假设到11月中旬，供应商明确需要达到每周生产2套的产出，就有足够时间建立相应能力，并配置人力来生产这些部件。

提问者：管理层的目标是在27年生产100台。从工程或供应链角度来看，您认为这一目标能否实现？根据您目前观察到的情况，较为合理的数量是多少？

回　答　者：这取决于采用什么基准。例如，蒙特利尔工厂每季度大约处理130个主要模块，我目前仅指气动相关模块。其中，每季度大约有40个是单独销售的独立模块，也可能略少，假设为30个，那么剩下就是100个主要模块。

这些模块大致相当于30台发动机。这里采用向下取整，约为每季度30台发动机。这个数字尚未计入迈阿密工厂和罗马设施，后两者也具备对部分发动机进行维修的能力。我对它们的具体数据不太熟悉，但估计迈阿密每季度约为15台，罗马每季度约为10台。暂且不将它们纳入计算，蒙特利尔每季度可处理30台发动机，也就是每年约120台。

从当前产出能力来看，这基本上已经是我们现有的处理规模。当然，维修与制造或改装在具体流程上有所不同。但考虑到目前已经具备这样的产出能力，而且公司正在采购若干发动机，用于建立拆解线，并利用不可用材料或剩余寿命较低的材料重建发动机，我认为100台是可以实现的。

不过需要说明的是，人们通常认为这些发动机将以全新零部件为基础进行生产，但实际情况并非完全如此。

回　答　者：大部分将通过改装完成，主要采用改装与供应链采购件或再制造件相结合的方式，最终组装成Mod 1。这个目标激进吗？是的。是否可行？我认为可行。唯一的挑战是建立一套严格的体系，以高效完成改装，同时还需要供应链提供支持，补足可修复零件方面的缺口。

提问者：您能否介绍一下，如何在传统航空业务与发电业务之间分配这些模块？传统航空业务是否享有优先权，其余模块再分配给发电业务？具体如何分配？

回　答　者：目前，蒙特利尔工厂有2条主要装配线，其中较新的一条显然用于发电业务。如果要估算发电业务需要分配的产能，坦率地说，由于人力和供应链等因素，其中一部分会占用航空业务设施的产能。

目前发电业务仍在爬坡阶段，很难准确估算具体分配比例。但我预计，未来6个月内，蒙特利尔工厂的业务分配可能会达到50/50。从业务角度看，这是合理的。现实而言，我们目前签订的发动机交付合同将享有优先权。总体而言，这些业务的利润率也更高，因此公司在确定业务优先级时，肯定会进行资源重新分配。

从航空业务角度看，一个有利因素是，部分送入车间进行所谓维修的发动机存在不可重复使用的部件，其中一些发动机最终实际上会成为账面上的沉淀资产。例如，一台航空发动机进入车间后，其前端或风扇框架仍非常可用，采购价格也较为合适，低压涡轮的情况同样如此，但核心机需要进行一定修复。通常，这类发动机会成为账面上的沉淀资产，而公司则会设法修复核心机或其中部分零件，再将其调配至其他用途。

发电业务的优势在于利润率通常更高，因此公司能够承受拆分核心机所产生的成本，并对其进行必要修复，使其恢复至可使用状态，之后还可以将其直接用于发电业务。因此，未来车间内可能会出现更多用于拆解的发动机。从航空业务的角度看，产出量可能不会保持不变，但投入量可能与此前持平，甚至更高，而这些投入将用于补充发电业务的需求。

提问者：下一个问题与测试有关。目前的测试硬件与客户最终实际收到的产品有多接近？是否仍有重大的设计事项尚未确定？是否仍可能进行重大设计变更？

回　答　者：抱歉，您具体是指测试方面的哪些内容？

提问者：例如，在现阶段，我认为你们正在测试一些完整的……

提问者：你们的目标是在26年第4季度（Q4）进行首次交付，对吗？目前是否仍有任何重新设计或重大设计变更，可能影响这一时间表？

回　答　者：据我所知，没有。目前设计已经得到验证，现阶段主要是在确定其运行包线。

对于Model 1的第1次迭代，我预计不会出现可能导致进度推迟的重大重新设计。您的问题应该是，是否可能发生某些情况，导致所有数据失效并需要重新进行设计。我们已经远远度过了这一阶段，目前正在完成最终定型，并为发动机确定可行的运行范围。

提问者：既然目前没有重大设计变更，现有完整设备中已有多少台接受了测试？这些设备在满负荷下大约运行了多少小时？单台设备最长的连续运行时间是多少？

回　答　者：关于具体设备数量，我不确定是否可以披露。我只能说，我们拥有多台原型机，并通过轮换方式开展多项测试。在部分原型机停机期间，我们会对其进行维护，确保各项表现符合仿真结果和设计团队的预期。

改变发动机的运行条件时，必然会出现一些意料之外的情况。我们正在对这些情况进行全面测试，确保没有遗漏。拆解和重新维护是开展此类验证的主要方式。

我曾深度参与其中1台原型机的测试。在我参与期间，最长的连续运行时间大约为4或5小时，但这段时间并非全部处于峰值负荷。测试需要覆盖不同的运行状态，同时也必须遵循发动机运行的最佳实践，即让发动机在特定工况下运行一定时间，随后降低负荷，并测试不同的运行包线。

提问者：换一种方式来说，你们对单台设备设定的目标运行时长是多少？目前已完成其中多少百分比？在向客户交付设备之前，设备需要在工厂内运行或测试多少小时？

回　答　者：抱歉，您刚才最后一部分的声音中断了。您想询问的具体时间是什么？

提问者：在向客户交付设备之前，需要在工厂内运行或测试多少小时？

回　答　者：就运行时长而言，我们近期的目标是对发动机进行耐久性测试。我们知道发动机可以长时间运行，航空应用部分也已经过测试，因此这并不是我们主要担忧的问题。不过，我们希望发动机能够连续运行1周，随后进行分析，确认发动机部分区域是否存在此前可能忽略的额外蠕变问题。我们重点关注LPT第2级叶片和HPT叶片，因为这些是需要重点考察的部件。下一个里程碑是确认发动机能够在预期工况下长时间运行。

提问者：关于热耗率和效率，完整机组目前测得的效率或热耗率是多少？能否同时达到输出和排放目标？

回　答　者：这个问题我无法提供完全准确的细节，因为我并未深入参与排放方面的工作。如前所述，我主要负责配置和硬件决策。关于热耗率和效率，目前发动机的表现符合预期。由于PTO轴、发电机和齿轮箱等部件的影响，确实存在一定的效率损失，但目前仍处于目标范围内。

关于排放目标，我无法提供过多信息，因为我对这方面的参与有限。我只能说明，目前燃烧室采用的是第1次迭代设计，也是CFM目前生产的最高效设计。我们计划继续对其进行迭代，以进一步降低排放。

提问者：您刚才提到，发动机已经进行了超过1周的全天候连续运行测试，对吗？在全天候连续运行过程中，哪些部件最容易出现磨损或故障？

回　答　者：并非如此。此前我已简要提到，根据模拟结果和目前的观察，部分部件更容易发生故障。这里所说的磨损和故障，是指超出CFM56发动机正常预期范围的情况，因为原始设备制造商既定包线范围内的其他情况已经纳入考虑。

此前未必能够预见的问题，主要是通过模拟分析发现的蠕变问题，因为蠕变是我们当前情况下导致故障的主要原因。低周疲劳也是一个方面，但通常可以依据LOP中的循环次数进行控制，因为热循环是导致低周疲劳的主要因素。除此之外，其他问题基本与高周疲劳或蠕变有关。

回　答　者：关于蠕变和高周疲劳，根据CFM的所有历史数据，我们尚未发现任何迹象表明部件会随机开裂或失效。为确保没有遗漏任何问题，我们正在重点检查LPT第2级叶片，因为这些叶片仍处于高温环境中。制造时，这些叶片并非采用单晶材料，但带有防止蠕变和扭转变形的涂层。然而，目前尚不能保证完全不会出现问题，因此我们正在确认这些部件能否达到全生命周期要求。

由于这些叶片并非受跟踪部件，为防止出现问题，我们正将其更换为新叶片，以此建立基准，明确新叶片的预期表现。我们也在寻找正在开发的替代方案，将叶片改为单晶材料，避免其像目前这样受到蠕变影响。此外，我们也在检查HPT叶片，以确保不会出现蠕变裂纹扩展。目前尚未发现此类问题，但这两类叶片是我们重点关注的2个主要部件。其他通常会因低周疲劳而失效的部件，其表现基本符合预期，我们暂时不认为会发生蠕变失效。

提问者：在测试过程中，哪类问题最需要进行故障排查？您能否举例说明出现了什么问题、如何修复，以及如何验证修复效果？

回　答　者：测试期间遇到的主要问题与VBV系统有关。在测试不同燃料和不同运行工况时，需要重新定义发动机的运行包线，例如加速或减速的速度，这涉及若干需要重新设计或重新评估的参数。

测试中的主要问题确实集中在VBV系统。发动机启动时，如果前端采用不同的设计，例如不同的增压级设计，进气流就会发生变化。进气流变化后，N2或核心机的响应也会不同，因此必须调整燃油供给。但如果燃油调整方式不当，例如供油过多，核心机就可能失速，因为VBV系统没有按照应有方式运行。又或者，为补偿快速启动而喷入过多燃油时，VBV系统会向ECU发送信号，提示核心机出现问题并要求停止运行。这属于较为复杂的PID系统问题，需要进一步分析并进行逆向排查。

这无疑是最主要的问题。其他因素基本都已纳入考虑范围，包括VBV和VSV，因为相应的调度逻辑也略有不同。如果能够控制好VBV，基本就能确定停机和启动的运行包线。此后进入Max Con工况时，实际变化并不大，主要是监测N1。

回　答　者：需要确保发动机没有超速，同时监控N2，避免出现过热。

提问者：VBV系统是由你们自主设计，还是来自供应商？

回　答　者：VBV系统仍采用OEM方案。严格从零部件角度看，VBV阀门本身全部是OEM零部件。我具体指的是其控制系统，本质上属于逻辑门。硬件保持不变，但由于进气系统不同，背后的控制逻辑也略有差异。

此外，除了增压器叶片外，进气导向叶片也存在细微差异，同样会影响气流，而OEM设计中并不存在这一结构。因此，测试时确实面临较大挑战。

提问者：为Model 1确定最佳VBV方案是否困难？还是经过几次测试后，就能找到合适的方案？

回　答　者：确实需要反复试验。刚才讨论的工作经历了约2.5周不间断测试，我们通过测试设计并迭代运行包线。每当发动机失速后，都需要执行一系列复杂程序，确保发动机能够再次正常运行，因为失速并非正常工况。

这项工作需要时间。每次超出或未达到运行包线，都会导致进度受阻。不过，通过持续测试和迭代，我们可以确定在特定加速率下允许的最大燃油输入量，并逐步完成整个瞬态阶段的验证。

提问者：这个问题不会影响时间表，对吗？你们仍将在4Q完成首次交付？

回　答　者：不会。我刚才所说的工作在数月前就已经完成。至少对当前这台原型机而言，应该不会受到影响。目前我们只是在微调运行包线。

我们已经明确了不可进入的区域，也就是运行极限，接下来需要继续迭代。可以从运行循环的角度理解：当N1从0 RPM提升至4,700 RPM时，这一瞬态过程中的加速度、速度变化或转速变化，都对应一组相互关联并持续变化的特定参数。有时系统受N1限制，有时受N2限制。真正困难的是准确界定这些条件，并围绕它们建立PID控制系统。

目前这一阶段已经完成，因此整体工作相对容易。我们现在只是进一步微调运行包线，以确保系统在效率基准或运行指引范围内运行。例如，燃油喷射量不应超过2,000磅/小时。

回　答　者：需要将其提高，否则容易出现超调，或者导致核心机堵塞或供给不足。

提问者：首个商业化机组需要采用什么样的排放控制配置？SCR 应用于哪个环节？目前是否已适用于该机组，还是后续升级的一部分？

回　答　者：正如我之前提到的，对于排放问题，我参与得不够深入，因此无法提供太多信息。排放控制的关键部分在于燃烧系统，可以采用双喷射、水冷等方案，这些方案目前都在考虑范围内。就现阶段而言，其排放水平与 CFM 技术插入型发动机最佳迭代版本的排放水平相近。其他方面我不太确定，因此无法进一步说明。

提问者：哪些部件、认证或其他因素可能会影响产品在目标市场的交付或运行？您是否了解需要取得哪些认证或批准？

回　答　者：主要因素可能是排放和噪声，包括 SFC 和噪声。当前的降噪效果较好，从认证角度看，我认为不会存在无法满足的要求。过去数月，我们一直在迈阿密以非常基础的原型机形式进行测试，目前已经基本满足该城市的噪声排放标准。如果能够满足城市标准，在美国西部等工业区域应该也可以满足要求。

噪声排放方面，我确定能够达标。对于 SFC 排放，由于我不太熟悉相关法规和认证要求，因此无法非常有把握地说明。唯一可以确认的是，如果 CFM56 技术插入型发动机的排放水平能够满足 SCR 的要求，那么该机组的排放将达到相同水平或更低。

提问者：客户在接受机组之前需要看到哪些验证结果？成功完成工厂测试是否足够，还是还需要在现场进行更长时间的运行测试？

回　答　者：如果客户采取非常保守的态度，可能会希望同时看到这2项测试结果。工厂测试目前基本已经完成，更长时间的现场运行测试也将很快开展，尤其是动力方面的测试；气动部分已经得到验证。

如果我是客户，在接受机组之前，我肯定希望看到工厂测试已经完成，并获得后端运行数据，以证明设备至少能够连续运行1周，并根据计算确认其可以连续运行1个月而无须停机。此外，我还会关注发动机监测和持续运行是否方便。我们也正在推进相关工作，以确保该机组易于集成。

提问者：我想了解客户侧的测试流程。对于现有客户，他们只需要查看测试报告，还是会派人到工厂现场见证测试？对于潜在客户，他们会如何验证和测试你们的模块及Mod 1发动机？

回　答　者：已有部分对产品感兴趣的客户参与了测试，其中也包括已经确认订购并为首批设备投入资金的客户。向客户展示严格且高效的测试程序，以及公司如何确保在规定时间内交付最小可行产品，能够显著增强客户的信任。这也可以让客户了解公司对各项参数的控制能力，以及确保产品交付的能力，因此非常重要。您问题的第2部分具体是指什么？

提问者：对于潜在客户，他们会如何见证测试，或者判断你们的产品是否值得信赖？

回　答　者：实际上，除了亲自查看产品及其背后的测试流程外，最终能够证明产品可靠性的只有执行情况和实际结果。在结果得到验证之前，客户无法完全依赖这些结果。对于刚进入市场的新产品，通常需要经过1年或2年的实际运行，才能充分判断其能否满足对应的细分市场需求。

就目前的市场情况而言，采购任何西门子发动机都至少需要2年的交付周期。如果客户当前迫切需要能源吞吐能力，主要有2种选择：第1种是选择成熟、可靠的西门子发动机，但需要等待2年；第2种是选择Mod 1，在未来几个月内获得设备。最坏的情况下，如果客户对其吞吐能力、支持服务或其他方面不满意，至少在未来1年半内仍有设备可用，之后仍可根据需要选择西门子。最理想的情况下，Mod 1在价格、效率和模块化程度方面远超预期，客户将获得一款具备竞争力且可以继续订购的产品。这是我对当前情况的看法。

提问者：在客户方面，你们此前披露了一项与埃隆·马斯克有关的重大合同。请问是否还有其他主要CSP，即云服务提供商？

提问者：主要的云服务提供商或AIDC开发商是否已经到贵公司的工厂看过测试？我认为，他们会考察包括Model 1在内的所有动力设备供应方案。请问，他们是否已经看过贵公司的产品或测试报告？

回　答　者：我目前不了解相关情况，因为我较少参与商业事务。如果您询问萨马穆德，他可能会详细介绍。我虽然与他共事，但我们更多讨论技术问题，很少涉及商业层面的情况。很抱歉，我无法回答这个问题。您可以查阅已经公开发布的报告。

提问者：从逻辑上看，如果客户采购西门子的航改燃机，需要等待2到3年，那么在未来1.5年内，Model 1是否是客户仅有的选择？市场上是否还有其他航改燃机供应商？

回　答　者：市场上有一些参与者正在起步。正如您刚才提到的，埃隆·马斯克曾表示，他们计划在内部开发航改燃机。如果具备库存以及现成的周转能力，也就是相应的产出能力，这种做法是合理的。

但我不认为埃隆·马斯克能够在未来1.5年或2年内迅速获得数1000个乃至数100万个航改产品零部件，再进行改造。我们的定位非常有利，因为我们已经拥有能够在CM56产品线内进行改造的产品，也了解这款发动机的局限性。此外，我们拥有3、4个测试设施，积累了大量可供参考的测试数据，因此对这款发动机有较深入的了解。

如果从零开始，我个人认为，限制因素不一定是硬件，而是与控制系统相关的一切。我一直在强调这一点。如果不了解发动机的运行特性，就很难从零开始开发一款新产品。我相信埃隆·马斯克最终可能会开发出来，但我没有参与其中，无法具体评价。不过，我认为他会发现，实际难度远高于预期。

提问者：您认为埃隆·马斯克会自行生产全部零部件，还是只生产叶片等供应最受限制的零部件？

回　答　者：您现在问的是我们的供应商吗？

提问者：不是。埃隆·马斯克表示将生产发动机，但我认为他不会独自生产整台发动机。他是否可能只生产叶片等供应最受限制的零部件，而其他零部件则采用外部供应？

提问者：这些都来自原始设备制造商，对吗？

回　答　者：是的。从FAA的角度来看，如果不是型号合格证持有人，就很难生产市场需求最旺盛的零部件。例如，目前需求最大的一些零部件是寿命限制件（LLP）和高压涡轮叶片（HPT叶片）等。就CFM56而言，除CFM和Chromalloy外，能够生产这些零部件的企业并不多。

Chromalloy刚刚开始生产高压涡轮叶片PMA零部件，但即使是他们，实际产量也未达到其宣称的水平。至于CFM，其产品仍有很长的等候名单。我们每隔3周或4周仍只能收到约30%的套件，具体周期我记不清了。这类能力无法简单地从零开始建立。

以寿命限制件为例，其制造流程极为严格。只有原始设备制造商生产这些零部件是有原因的，因为他们掌握专有工艺和相关最佳实践。尝试改变制造工艺会产生明显后果。例如，我记得可能是Pratt，也可能是Rolls；我认为是Pratt开始采用粉末冶金工艺而非传统铸造工艺生产寿命限制件，但其零部件寿命并未达到预期标准。一旦改变制造方式，情况就会变得复杂，因为制造工艺会直接影响零部件的使用寿命。

如果他们能够生产叶片，这当然很好，但我不确定这与我们目前合作的其他PMA制造商相比有何差异，除非其生产的是高压涡轮叶片和寿命限制件等需求极高的零部件。作为非型号合格证持有人，关键问题在于如何证明产品符合持续适航指令（ICA）的要求。实际上，企业必须向FAA证明自身具备相当于原始设备制造商的能力，这会成为一个棘手的问题。

提问者：总体而言，他们要在2至3年内完成这项工作，将面临很大挑战，对吗？

回　答　者：我认为确实如此。仅以高压涡轮叶片为例，Chromalloy大约用了4或5年才获得FAA批准。正如我所说，如果是其他类型的翼型件，市场上已经有大量供应商能够生产PMA零部件。针对动力业务，我们已经在与这些供应商接洽和合作。

Pacific Aero就是其中之一。我们收购了该公司，因为它具备很强的叶片生产能力。因此，在翼型件方面，我们已经具备充足能力。至于其他企业，我不认为其进入市场会彻底改变现有格局，因为最终只会成为另一家PMA零部件供应商。目前真正存在供应问题的是高压涡轮叶片和寿命限制件，而我不认为寿命限制件能够突然由第三方制造。

提问者：再谈谈新叶片的供应链。对于Model 1，你们需要用新叶片替换旧叶片，对吗？目前这些叶片有多少家供应商？他们是否已经准备好为你们扩大产能？

提问者：关于27的产能爬坡，想进一步确认相关叶片的更换情况。

回　答　者：需要澄清的是，我们并不会将所有叶片都更换为新品。您指的是风扇叶片吗？

提问者：是的。

回　答　者：风扇叶片将采用新制造的产品，供应仍由会议开始时提到的供应商负责。供应商已承诺在11月中旬达到既定产量，我们目前也已经拿到了一些产品。因此，风扇叶片和增压级叶片都没有问题。

另一个提到的部件是低压涡轮第2级叶片。目前采用的是新品，但这不会长期作为OEM标准方案。我们将采用单晶叶片这一替代路线，该方案正在推进中，但目前还没有明确的完成期限。

现阶段，新低压涡轮第2级叶片的采购并不是太大的问题，因为可以直接从OEM购买，而且与发动机的其他部件相比，这类叶片的需求并不高。

提问者：大约2至3个月前，业内曾担心叶片供应链问题，认为叶片会成为转换至Model 1的主要制约因素。您认为目前这已经不再是主要制约因素，对吗？

回　答　者：并不完全是制约因素，因为在翼型件方面，没有任何因素阻止我们使用PMA零件，也没有任何因素阻止我们进行内部维修。

从航空公司的角度来看，例如American Airlines通常不会在发动机上使用任何PMA叶片，无论是压气机叶片、低压涡轮叶片、高压涡轮叶片，还是增压级叶片。因此，它们只能进入OEM的等待名单。例如，需要等待1个月才能获得3套叶片，那么每个月最多只能完成3台发动机。在这种情况下，叶片供应确实会成为一个重大问题。

但在我们看来，一方面可以依靠OEM。我们与OEM签有MCPS协议，既能以较好的价格采购OEM零件，也能在其释放的部分叶片套件中获得一定的优先权。另一方面，我们还可以依靠PMA供应商，包括Pacific Aero、Chromalloy和Heico。我们合作的供应商很多，因此可以从不同供应商处采购。

我们具备使用PMA零件的灵活性，而对于航空领域的大型航空公司，由于其维修规范已经确定，使用PMA零件会更加困难。因此，我认为翼型件方面的供应基本已经落实。可能需要关注的是风扇叶片，但相关供应将在11月中旬到位，因此应该不会有问题。

此外，我们还将更换低压涡轮第2级叶片。不过，这类叶片目前需求不高，因此直接从OEM采购新品并不困难。

回　答　者：从计算结果和航空应用的角度来看，我认为实际情况并不像人们直觉上认为的那样糟糕。更具挑战性的可能是那些不容易被直接观察到的部分。对于能够看到的问题，人们较容易判断哪些因素会形成瓶颈；但对于看不到的问题，则更难预测并采取相应的补偿措施。

提问者：我们来讨论一下经济性。目前，可以改造成 Model 1 的二手 CFM56 发动机采购价格是多少？此外，完成 Model 1 改造所需采购的其他必要部件，成本大约是多少？您对此是否了解？

回　答　者：我不清楚整套封装系统的具体成本，因此只能说明发动机本身的情况，不包括额外配置的 PTO 轴、发电机和齿轮箱。我估计这些设备可能需要额外增加约 2百万美元，但并不确定。

对于 CFM56 发动机而言，价格取决于其剩余寿命和核心机的状况，这些是决定账面价值的主要因素。一般来说，一台非常适合发电应用的发动机，前期采购成本可能为 4百万美元；核心机维修和修复成本约为 3百万美元，主要是因为需要更换或修复大部分已经使用过的叶片，并对核心机进行全面检修，同时保留限寿件等部件。这样计算下来，成本约为 7百万美元。

如果再增加 1百万美元用于机组、附件和 QEC，成本将达到 8百万美元。再加上前述封装系统的前期成本，我暂按 2百万美元估算，尽管不确定实际成本是否确为 2百万美元，总成本大约为 10百万美元。

一般来说，一台全新的发动机——准确地说，是一台循环次数较高、因而在航空应用中更具价值的发动机——如果状况良好，售价约为 12百万美元，通常在 10百万美元至 12百万美元之间。

对于发电应用，发动机不一定需要具备较多循环次数，因为更关注的不是低周疲劳，而是蠕变。因此，与航空应用相比，可以采购价格更低的发动机，同时获得更高的利润率。换言之，前期投入成本可能与航空发动机核心机全面修复的成本相近，但利润率更高，因此可以在修复环节投入更多，以确保发动机能够达到所承诺的使用寿命。

提问者：如果使用这些 PMA 部件，采购价格大约可以降低百分之多少？

回　答　者：这是一个很好的问题。对于主要部件，我认为无法披露具体数据。不过，可以以机组为例进行说明，因为我最近参与了相关工作。

回　答　者：在制定维护方案时，以VBV系统中的滚珠丝杠作动器为例，OEM认证翻修件与PMA翻修件的价格差异很大。假设OEM翻修件的价格为$10,000，PMA翻修件则约为$2,000至$3,000。对于航空发动机中价格较高的部件，例如HPT叶片，成本差距虽然没有这么悬殊，但仍然非常显著。如果一整套部件的价格为$2 million，采用PMA翻修件可以大幅降低成本。以HPT叶片为例，每套可节省接近$1 million。对于LLP，我们不会偏离原有标准。HPT NGV也是如此，整套价格非常昂贵。从单件来看，我认为它们是发动机中最昂贵的部件。

回　答　者：这类部件可以获得40%至50%的价格折扣。

提问者：这个价格确实很有吸引力。下一个问题是，Model One的实际单机输出功率是多少？管理层公布的是25 megawatt，但实现大规模生产后，每台设备的实际输出功率是多少？

回　答　者：接近25 megawatt，在20 megawatt以上。

提问者：具体情况如何？

回　答　者：目前有些信息还不能全部披露，但就第1台原型机而言，在燃烧室尚未进行任何改进、VBV系统调度也未得到准确或充分优化的情况下，我们已经实现了21 megawatts的总输出功率。因此，达到25 megawatts是可以预期的。

此外，还需要考虑到设备目前在迈阿密运行，当地可能具备最不利的运行条件，即高湿度和高温。任何优于这一条件的环境，设备表现都应当更好，尤其是在EGT决定输出功率的情况下。

例如，如果1月在蒙特利尔进行测试，室外温度为minus 15，最大连续运行状态下的EGT约为800 degrees；如果是在夏季，EGT则会达到850。因此，EGT裕度基本上接近one to one的相关关系。一旦进入更干燥或温度较低的环境，发动机有望输出更多功率，因为在EGT相同的情况下，可以让发动机以更高转速或更高负荷运行。

提问者：目标输出功率是多少？是22 megawatts还是23 megawatts？

回　答　者：目标输出功率至少为25 megawatts。我认为25 megawatts是非常合理的目标。真正决定产品竞争力的因素，将是排放水平、达到25 megawatts所需的燃料量，以及发动机实现这一输出功率时的效率。

提问者：关于测试，你们已经进行了整套设备测试，还是只进行了燃烧室测试？

回　答　者：您所说的整套设备测试具体是指什么？

提问者：这是完整的整机测试，还是仅测试发动机？

回　答　者：这是完整的整机测试，涵盖整台发动机、PTO 轴、齿轮箱和发电机。

提问者：关于风险，假设首次交付或产能爬坡推迟6个月，最可能的原因是什么？

回　答　者：从最坏的情况来看，可能是出现了此前未预见的某种蠕变失效，尽管我认为这种情况发生的可能性较低。这类问题可能导致我们长时间反复迭代，也是少数可能造成项目推迟6个月的情况之一。6个月是非常严重的延期，意味着需要进行重大重新设计，尤其是考虑到我们目前所处的阶段。因此，必须是此前未预见的蠕变失效，才可能造成如此严重的延期。

更可能出现的情况是燃烧室在排放方面存在设计约束，这可能是最容易导致延期的因素。但我认为这不会使整个生产计划推迟6个月。正如我之前所说，Mod 1有不同的迭代版本。大家所理解的版本，可能相当于我所说的第2次迭代。第1次迭代不完全是原型机，而是用于概念验证，证明航空衍生型发动机能够正常运行并达到25兆瓦。第2次和第3次迭代的主要目标，是进一步提高整体效率并降低排放。

因此，燃烧室设计最有可能导致进度推迟，但不会使整个生产计划推迟6个月，更可能只是使第2次或第3次迭代推迟几个月，因为相关工作需要更多时间。

提问者：最后1个问题，FTI的工作在哪里结束，Jera的工作又从哪里开始？你们与Jera设有合资企业，对吗？最终系统测试和问题修复由谁负责？

回　答　者：目前，发动机的生产或组装工作在蒙特利尔进行，这部分由FTI负责。随后，发动机会被运至迈阿密，与PTO、发电机和齿轮箱一起完成测试及后续工作。

之后，Jera会将这些设备集成并封装成新闻报道中提到的移动式拖车发动机系统。Jera主要负责最终封装。因此，FTI的工作在除发动机外部壳体之外的所有部分完成后结束，外部壳体就是拖车式结构。具体而言，当发动机完成全部组装，并与PTO一起交付至Jera的设施后，FTI的工作即告结束。

提问者：发电机和齿轮箱由谁负责封装？最终系统测试由谁负责？

回　答　者：最终系统测试也由FTI负责。整体流程是在蒙特利尔制造，在迈阿密测试，然后运送至Jera进行封装，之后再运回FTI完成最终测试。

提问者：对于Jera，如果目标是在27年达到100台，其产能是否足以支持装配线爬坡？您认为Jera在爬坡过程中是否会面临限制？

回　答　者：我对封装环节的参与程度不如其他人员深入，但据我了解，目前一切进展顺利。当前主要问题是准备驱动装配线所需的材料，但就现阶段而言，该工厂的能力符合我们的标准。

封装工作尚未开始，因此Jera仍有时间进行调整。据我了解，目前整体情况基本符合预期。Jera在调整封装流程方面效率较高，能力也处于较好水平。

提问者：近期Crusoe放弃了Boom的Superpower订单，您对此怎么看？FTI同样是固定式发电领域的新进入者，Model 1面临哪些相似或不同的风险？

回　答　者：这件事确实是近期发生的，但我不确定Crusoe选择退出的具体原因。据我了解，他们希望获得的方案可能不只是涡轮机或航空衍生型燃气轮机。我也不清楚该项目及订单合同的全部细节。

如果Crusoe原本期望在明年或明年年中获得产品，而Boom Supersonic表示最早只能在2028年初交付，那么交付时间可能是影响因素。总体而言，项目推进和交付速度非常重要。如果交付速度不符合客户预期，客户通常会更快寻找所谓的更优选择，这会影响其决策。

从航空衍生型燃气轮机的角度看，我目前想不到有其他公司能够达到我们的交付速度，这可能是我们的主要优势。客户选择退出的原因未必是时间，也可能与效率或现场支持有关。

回　答　者：从灵活性的角度看，如果出现一种成本更低的全新氢能发电方式，这是我目前唯一能想到的风险。

提问者：您认为这会构成重大风险吗？例如，可能有新的氢能企业进入市场？

回　答　者：我之所以提到这一点，是因为我的1位好友正在从事相关工作。我认为氢能的能量密度尚未达到同等水平，只是在考虑一种绿色、能够发电且易于获取的能源时，首先想到了氢能。我不认为它目前处于同等水平。

未来可能会采用多种能源组合。例如，一家大型承包商需要能源供应时，如果预算足够高，可能不会完全依赖航改型燃机，而是根据实际情况，将太阳能、电池等多种方式结合起来，甚至也可以与煤炭混合使用。

归根结底，能源需求方并不一定局限于某1种特定的供能方式。他们需要的是能够快速、高效且价格合理地获得能源，具体采用何种最终产品并不重要，关键是能源供应必须可靠、能够尽快交付，并且价格尽可能低。

提问者：最后1个问题与传统航空AP业务有关。就PMA零部件而言，客户对PMA零部件的接受度呈现怎样的趋势？目前AP业务的利润率面临压力，使用PMA零部件可以提高利润率，但客户是否接受？原厂零部件企业采取了哪些措施来抵制PMA零部件？

回　答　者：这个问题本身值得单独深入讨论。先回答第1部分。就目前的PMA市场而言，特别是CFM56发动机市场，当前可能处于历史上最好的阶段，因为这款发动机正在逐步接近生命周期末期。此后已经有7B、LEAP以及其他几款已发布的发动机设计，正在逐步取代现有机队。

在继续运营这些发动机时，客户需要具有成本效益的维修计划、项目和替代方案。由于OEM正在逐步停止相关生产线，或减少OEM零部件的供应，客户不得不寻找其他替代方案。

过去1年，我看到航空公司的态度发生了非常明显的变化。部分航空公司过去从不提及PMA，甚至完全不会将其视为可选方案，但现在已经开始明确选择配备PMA零部件的发动机。具体信息无法披露，但可以确认，有1家航空公司购买了我们的发动机，其中采用了全套PMA零部件，包括HPT NGV、HPT叶片、LPT第1级NGV和HPT护环。

回　答　者：从护罩、HPT压气机叶片到其他部件，全部采用了PMA件。他们采购后从未对外宣传，但业内人士都知道，也都在关注搭载这台发动机飞行的飞机，并将其视为机队中的先导样本。他们希望了解这台发动机能否像机队中的其他发动机一样可靠。如果答案是肯定的，我相信他们会对PMA件更感兴趣。

如果在5年前询问业内人士，他们可能会表示绝不会在发动机中使用PMA件，理由可能是其尚未得到验证，或存在其他顾虑。但目前，很多人虽然不一定会选择PMA产品中最昂贵的部件，却开始考虑重要性相对较低的部件，例如起动机。易捷航空需要补充50台甚至60台发动机起动机，但目前陷入困境。供应链已经停滞，他们需要等待霍尼韦尔供货，却始终无法获得产品，因此只能转向PMA件。这已经成为当前行业的普遍做法。

目前PMA件的发展状况可能是有史以来最好的。就替代方案而言，我认为对PMA的负面印象主要存在于航空航天行业。在工业领域或航空衍生型发动机领域，这种看法并没有那么负面，因为相关设备并不用于飞行。运营商或航空公司在飞行中使用PMA件时，最主要的顾虑是安全问题。我认为这种顾虑缺乏依据，不过也有人持不同意见。在工业航空衍生型发动机领域，PMA件是一个更具吸引力的选择，吸引力要大得多。

提问者：好的，非常清楚。今天的访谈就到这里，祝您今天愉快。谢谢，再见。

回　答　者：谢谢，祝您一切顺利。

原文转译

And what could slow down delivery or the production ramp? Okay. Okay, the the the first question is, um, could you share how closely have you been have you been involved in Mod 1? Uh, which parts of the engineering or testing work have you worked on directly?

Um, yeah, I mean, I've been involved uh directly in multiple facets. Uh, I've been involved in the configuration portion of the engine to make sure that from a configuration perspective it works. Uh, I've been involved in the design phase. I've been involved in the testing phase as well that's still ongoing. Um, so broadly speaking, I've I've seen all the facets, maybe not the uh the marketing aspect and the packaging portion as much, but Mhm. Yeah.

Okay, sure. Uh, and then then uh, the next question is, where does Mod 1 stand today? It's uh engine testing or it's complete generator set testing or operation at uh at some site? Uh, which which stage it is in?

Um, technically it's still in the regulation phase of testing. Um, I think that's all I can say. I mean, most of the um, the public statements from FTI still stand to this day. Um, the first iteration of Mod 1 is currently being built in the Montreal facility, but the testing is still ongoing um at the Medley site in uh Florida.

Mhm. Okay, okay. And the management has talked about a first delivery in Q4 26, right? Uh, what does first delivery mean in practice? It means uh operation at customer site? Uh, it means uh shipping or commissioning or or or or get customer acceptance?

Uh, yeah, that's a good question. Um, so the first uh delivery means um the engine being operational, not necessarily fully packaged and delivered uh from what I can from what I understand. So in between the customer acceptance and um being fully ready, operational and delivered to the customer site.

Mhm, mhm. Okay. And um could you tell us like what still has to happen before the first unit is ready? So which critical steps really sets the timetable?

Sure. Um, so in order to get the first unit ready, uh a lot of things need to to be done. Uh as you know, this is a aeroderivative engine. Um, so therefore there needs to be a proof of concept for everything that is different within the engine, namely for Mod 1 specifically, uh the booster and everything that has to do with the N1 portion of the engine is um uh a little different, I'll say. Um, a portion of the combustion section as well is uh quite a bit different. So those um still need to be uh proof tested. Um, the operational envelope of the engine has to be uh defined or finished, I should say. Um, and from that operational envelope, then some some control systems need to be established to make sure that the engine can run within the uh envelope that we want it operated under. Um, in terms of steps that set the timetable for like the first few units, this is mainly it. Uh, when you think about long-term sustainability, um, I think the supply chain uh as well as a rigorous and, you know, flexible, I should say, um, throughput uh of material to convert, you know, aero engines into aeroderivatives is going to be the the the step that makes it um, how can I say this? Like this is going to dictate essentially the the delivery of the engines. But for the first one, it's going to be the operational envelope.

Mhm, mhm. Okay. Uh, let's put it this way. So could you walk Walk me through what changes um when you turn a CFM56 into a Mod 1. Uh which parts get reused, um modified or newly designed?

Yep. Um I've kind of briefly touched on this, but um the main portions of a CFM56 get reused. So I I think I can get into a bit more of the the details if you'd like. Um but essentially everything around the core except maybe the combustion section and some little uh tweaks um are essentially the same. So what you need to turn a aero engine into a Mod 1 is essentially 90% of a uh CFM56 engine. Your LPT for the first iteration of the Mod 1 is relatively unchanged. Um you need to redesign the forward end which I mean it was unveiled I I believe through a um investor's call uh but and I kind of briefly talked uh about it but your forward end of your engine gets redesigned. So when you think fan blades, they're they're different. Uh the booster section there's a there's a an inlet guide vane that uh gets added in front of it to kind of regulate the airflow. Um everything that's bypass flow gets essentially nullified. Um additional instrumentation gets added to it as well as uh other harnesses that are normally used for the nacelle and uh the accessories of the aircraft can be removed. Um so the yeah the big three points are your your booster, your fan blades, your combustion section and uh everything that has to do with the uh the control system.

Mhm mhm. And uh of those materials, so how how much of them in Mod 1, so how much of them uh uh would use the proven hardware and how much needs new development and testing?

Yeah uh okay. So how much of it gets reused was the first question, right?

Uh yes. Or it's like uh when you design the Mod 1, uh you you need to make some change, right? Uh redesign. So I I want I want to know that how how much of the of the materials or or or parts you could use the reliable uh uh supply chain or some or how how how much you need a new development and you maybe bring in new supply chain so which I I think is time consuming, right? So so I I try to get uh I want to figure out uh how much of them, how much of the modification or how much of the uh new uh new parts uh you could use reuse the the the proven hardware or proven parts and uh how many how how much of them you need to redesign and and test the new parts.

Yeah no okay okay I understand. So I mean kind of like I I've discussed like the probably like 90% of a CFM56 engine can be reused. So from when it comes to a maintenance perspective, um you can essentially just copy paste the guidelines of the OEM. You can deviate on some portions based on an engineering perspective for for stuff that let's say doesn't need to be airworthy. I'll give an example. Um your bypass uh flow goes through OGVs. Those OGVs don't need to be serviceable for a power engine because you're not using the bypass flow to generate thrust. So um when it comes to to little parts like that, uh you know there's there's some um how can I say like re reiteration of the um the serviceability criteria that get uh to be touched on. Um but generally speaking 90% of the engine is the same. When it comes to the new parts uh or the redesigned portion, um namely I I think the the main one uh that required the most um I guess like uh input to to design and and to prove The uh it's it's a utility was the the fan disc. The fan disc is redesigned. Um normally that's a big uh it's a big step because technically speaking for an aero product uh this is a life limited part. Uh so rigorous testing and uh and uh analysis and computations goes in there to to make sure that you're able to reach um your threshold. Um in this case we're mostly concerned about creep. So the design specs is a is a bit different um but we're allowed to kind of go overkill when it comes to to design. So that gives us a bit of flexibility. So fan disc is one of them. Uh I briefly mentioned as well the fan blades. Those get redesigned um and then remanufactured. The I've mentioned as well the um inlet guide vanes for that booster get redesigned. There's another shroud that goes around those. Um um and the stage one booster blades are also redesigned. Well technically it's stage two. Um so those like the front end of the engine is mostly the the the one that takes the the bulk in terms of of redesign portion and remanufacturing processes that need to be uh taken into consideration. Uh and then at on the back end or the hot section it's also the uh the fuel nozzles um for now. Um then later on it's going to be the fuel nozzles, the fuel manifolds and uh potentially your um combustion liner because there's going to be multiple iterations of uh or configurations I should say of uh the mod one. So booster and combustion section are the the two main ones.

Mhm. Mhm. Okay. Uh for those re- redesigned parts, so are are there any constraints? Uh what what what about the lead time for for those parts?

Oh okay yeah no that's a good question. I mean we we are working with uh essentially like third party manufacturers that are able to to design those parts and for um sorry sorry for um let's say for example the fan disc I mean we have already some on site. We have multiple units that are um manufactured. Those parts were designed I believe um we're talking like six months ago. Um the manufacturing time is the lead time for each is about two weeks. Um so essentially whenever a order is placed the the moment the uh third party uh vendor essentially has all the drawings, all the specs uh we could essentially get it in like yeah two three weeks. Um and they can the throughput per week is about like two um two LPs per week. So if that answers your question that's yeah.

Mhm. Um so so but uh we talk about two to three weeks uh lead time. Uh what about the amount? Because when you ramp up in in in in the next year, right? Uh is is still the lead time is only as short as two to three weeks? Because before I thought like for example for for the fan fan blades uh uh uh the lead time may be two to three months. But now you are talking about two to three weeks. So so so uh so I'm a little bit confused. Oh I'm I'm so I'm thinking uh from what you say uh it seems that there's there's not as many constraints as I have thought in the supply chain.

No I see what you mean. That's a good point. I mean so right now again it depends who you ask and based on which state of the project you ask them about but uh I'm specifically talking about like the first iteration of the mod one. Um because there's multiple configurations. Um when it comes to the fan disc like I mentioned we already have a few units on site. I'm I'm from the Montreal shop. Um sorry so just want to point that out. So like um those parts are already on site and I'm working with them on the daily. Um, there's one that came back from the Miami testing facility last week, um, on the engine that was tested there. Um, and the new design of the fan disk is already on site. Uh, when you're talking about the, um, booster blades or the fan blades in question, uh, those are already on site as well. We have a, uh, we have a a little bit less iterations because we don't need them as much because we're still testing on the aero config portion, um, in Montreal. Uh, so we have two sets of those. The lead time on those after we finish, sorry, the, uh, the testing, I mean, the idea is to get them for November, so maybe that's when the, the dates that you've heard probably match. Um, but the idea is the moment you have a finalized design and you can prove that your testing envelope is correct, you can essentially just start to, um, ramp up the production and get units, um, within normally like, yeah, two, three weeks if the provider already kind of expects a, a repeated order, right? So by, let's say, I'm going to say a number, don't quote me on this, but let's say mid-November, um, the provider knows that they have to throughput two units per, per week, let's say. Uh, then it gives them enough to get the capability going, um, and the, the manpower allocated to essentially just manufacture those, uh, those parts.

Mhm, mhm, yeah. And, uh, and the management has, uh, targeted like one, 100, right, units in, in 27. Uh, uh, but from, from your point of view, uh, um, or, or from the engineering side or from the supply chain side, uh, do you think it's real, uh, realizable or, or, or it's like, um, uh, what, what would be the fair number from, from what you're seeing?

Yeah, um, I mean, it depends what, um, your basis is, right? Let's say, for example, the Montreal shop does around 130 modules, major modules per, per quarter. So that's, and I'm talking just aero right now, right? And out of those, there's maybe, I don't know, maybe like, uh, 40 that are standalone modules that we sell, uh, per quarter, maybe a bit less, let's say like 30. So now you're down to 100 major modules. Out of these, um, it's about an equivalent of, uh, of 30 engines. I'm going to round it down, but it's about 30 engines per, per quarter. Um, and that doesn't take into consideration the Miami shop and the Rome facility, uh, which also are able to do, um, some maintenance on some engines. I'm not super familiar with their numbers. I would assume that Miami is probably around the 15 mark, uh, and Rome is probably around the 10 mark per, uh, per quarter. But let's take them out of the equation. So you have 30 engines, um, per quarter. So that's about 120 per year. That's for Montreal. And so right now, like in terms of throughput, it's pretty much already what we do. Obviously, it's, uh, from a maintenance perspective, it's a bit different than from a manufacturing perspective or conversion perspective. Um, so if you take into consideration that you're already kind of doing that as throughput and that the company is currently buying, um, a few engines to kind of have like a strip line and rebuild, uh, engines from unserviceable material or material that is, uh, a bit too low on life, then I think that's an achievable number. But the thing that I'll say is that people tend to think that it's going to be, um, based off of new parts, uh, which is not exactly the case. most of them will be a a conversion and most of them are going to be like a mix of that conversion plus some um purchased or remanufactured parts uh from supply chain that will essentially combine together to create that mod one. Mhm. So, is it aggressive? Yes. Is it possible? I think so. Um the only challenge will be being able to have a rigorous system that allows you to do conversion efficiently. Um and the supply chain to kind of back it up uh to backfill the holes that you have in your uh parts that you're able to salvage. Mhm. Okay. Uh and uh could you could you walk me through that how how how you distribute the modules between uh the traditional AP power business and the power business? So uh is it a priority for for for traditional uh aerospace business uh uh and you take the left to the power or or how how you distribute the the modules? Yeah, uh I mean in the Montreal facility right now we have two main assembly lines. Uh the recent one is obviously power. Um if I had to guesstimate the amount that you have to allocate for power because uh for being honest, some of it will be kind of uh eating away at at the aero facility capabilities simply due to like, you know, manpower um and supply chain uh from a from from those perspectives, right? Uh how much you allocate, um I mean right now it's ramping up, it's kind of hard to to estimate. Um but my guess would be within the next six months Montreal will maybe be doing 50/50. Uh simply from a um like business perspective, right? Realistically speaking right now the contract that we have to deliver those engines will kind of take priority. Um generally speaking as well those uh those profit margins are higher. Um so as a business I'm sure there's going to be reallocation when it comes to uh what takes priority. And when it comes to what you can do from an aero side, the good thing is some engines that you get in the shop that you quote unquote do a maintenance on, some of it is not reusable and some of it kind of essentially kind of dies on the books, right? Let's say for example you have a uh an aero engine that comes in, um the forward end or the fan frame is super viable, um you get it for a good price, your LPT is the same, but your core is where it needs a bit of love. Um generally speaking those um engines tend to kind of die on the books and you're trying to find a way to salvage the core or some parts of the core to then allocate somewhere else. But what's good about uh about power is that because the profit margins tend to be higher, you're kind of able to take the hit from splitting that core up and doing whatever restoration you need to do on it to kind of bring it back to a serviceable uh standpoint. And then you're also able to take kind of just use it on power uh for that. So I'm sure you'll see a lot more of uh teardown engines uh coming through the shop. Uh so you might not see the same output from an aero perspective, but I think the input might be the same if not higher. But that's going to compensate to backfill power. Mhm. Okay. Okay, got it, got it. Um okay, the the the next question is like um about the testing, okay? So how how close the test hardware to what customer will actually receive? So are there any major design still open? So are there any major design change still open? What do you mean from from like a a testing perspective, sorry? Yeah, it's like uh at this point uh you are I I think assume that you are testing some complete Uh, I think I can can pre set. So, um, are there, uh, and you are targeting, uh, first delivery in Q4 26, right? Uh, in in the fourth Q, uh, do you still have any, uh, redesign or do you have still any major design changes which would, uh, affect the timetable?

Um, not that I'm aware of. Um, right now we're, like I said, we're, the design is proven. What we're doing right now is kind of just, um, kind of tracing that, uh, operating envelope, right?

Mhm. Mhm.

So in terms of major redesigns for the first iteration of model one, no, I don't foresee any major redesigns that could, um, push it back for, let let's say you're, I'm guessing what you're asking is essentially is there something that happens that, you know, could throw off all the numbers and we have to reiterate the, uh, the design process. But I think we're, I think we're well past that, uh, that curve here. I think we're, we're in the, we're trying to finalize it and establish a, uh, uh, operating, uh, regime that's viable for the engine.

Mhm. Mhm. Mhm. Okay, okay. I I get it. I I get what you mean. Okay. And, okay, uh, so if there's no major design, redesign, you well though. So for for the current, uh, complete units, so how many units have been tested and roughly how many hours have they run at full load? And, uh, and for one unit, what's the longest uninterrupted run?

Yeah, good question. Um, how many units, I'm not sure if I'm allowed to disclose this. Um, all I can say is that we have multiple, uh, prototypes that we use on on rotation to to test multiple things on. And in the dead time, we do some maintenance on the rest to make sure that everything is up to, uh, what is expected from the simulations and the design, um, team.

Mhm.

So obviously when you do change the operating conditions of an engine, you have unexpected things that happen. We're testing all of that and we're making sure that we're not missing anything. Um, stripping and reservicing is the main way to do that. Um, and so far the, I'm not sure if I'm able to disclose it, but I I've been closely, uh, involved in the test for one of those prototypes. Um, and when I was there, the longest uninterrupted run, uh, was probably on the upwards of like four or five hours. Um, although those, you know, those numbers were not necessarily under all peak loads. Uh, you're trying to test different, uh, different regimes, of course, and you're also kind of bound to the best practices of, you know, running the engine to a certain amount for a certain amount of time, um, and then kind of throttling back and testing different, uh, different envelopes. So I'm not sure if that answers your question, but that's what it is.

Mhm. Mhm. Okay. Uh, so, uh, I I let's let's put it this way. So what what is target the the the the hours, uh, you would run on on on on one unit and, uh, how much percentage you have, uh, completed? So be before you you deliver a unit to the customer, uh, what is required, uh, the the the test time, uh, on your site? You you should...

So before you deliver a unit to a customer, sorry, you kind of cut out at the end. What is the what time?

What what is the time, uh, you need, what how how many hours you need to run it or how many hours you need to test it, uh, uh, in in in your factory? Oh, I see what you mean. Yeah, okay. Um, in terms of hours, what we're, uh, aiming to do essentially, um, soon, I'll say soon, is, uh, to be able to do an endurance run, uh, on the engine just to see, because we know it can run for a prolonged amount of times. Uh, it was tested on the aero portion. So this is not really of our concern, but, uh, we want to say that we can run the engine for, uh, a week straight, um, to then dictate and analyze to see if there's any additional creep that we're, uh, maybe overlooking, uh, in some portions of the engines. Uh, specifically, we're not, um, targeting anything but the LPT stage two blades, um, and obviously your HPT blades, because those are the, um, the parts of, uh, of interest here. But this, this would be, um, the next, uh, the next milestone, I would say, to confirm that the engine can run under a desired, you know, uh, condition for a prolonged amount of times.

Mhm. Mhm. Okay. Uh, and the next question is about the, the heat rate, the efficiency. So what efficiency or heat rate has been measured on the complete package and can it meet the output and emissions target at the same time?

Okay, so this question I don't think I can fully answer with, uh, accurate details because I'm not really involved on the emission side of things. Uh, like I mentioned, I'm more involved on the configuration side of things and, um, hardware decision. But, um, when it comes to the heat rate and efficiency, um, what I can say is that the engine is, well, it's, it's reacting in a way that's expected. Obviously, there's, um, some, um, some efficiency losses due to, like, the PTO shaft, the, the generator, uh, and the gearboxes. But so far, we're within targets. When it comes to emission targets, um, I'm, I can't really speak too much on that. I'm not, um, as involved, um, on that side. All I can say is that the current design for the combustion chamber is the first iteration. It is the most efficient one that CFM is producing at the moment, and we're planning on iterating on it to, to make sure the emissions get lower.

Mhm. Okay, okay. Uh, I see. And, and the next question is about, um, about, you, you, you said you have already, uh, tested the engines for, uh, for, for a round-the-clock operation for more than one week, right? So, uh, I, I, I want to see, uh, which parts are most exposed to wear or failure in, in, in, in the, in the round-the-clock, uh, operation?

Right. No, um, I, I think, uh, I've quickly touched on that, but the parts based on our simulations and what we see, uh, that are more, um, exposed to failure, and when we say wear and failure, um, I, I guess we're, we're talking outside of what is expected from a CFM56 engine, um, because anything else that is already within the OEM envelope, we kind of take into, uh, into consideration. What we're not, um, uh, sorry, what was not, um, foreseen necessarily are everything that we ran through simulations for, um, creep, uh, because that's the main, uh, leading cause of failure in our case, um, because low cycle fatigue is one aspect, but that's generally controlled through your cycle counts from your LOP because of your, um, your, your heat cycles. Those are the main ones that, uh, tend to cause, uh, low cycle fatigue. And then everything else is essentially high cycle fatigue or, or related to creep. creep. High cycle fatigue wise, we didn't really uh see anything um from all the historical data from from CFM that could lead us to think that a part would randomly crack or fail. Um so in order to make sure that we're not overlooking anything, we're looking specifically at um the LPT stage two blades because those blades are still in a high heat environment. They are not made of a single grain crystal um when they're manufactured and they do have some coating that prevents them from creeping and untwisting. Mhm. Uh however, this is not um bulletproof at the moment. So we're making sure that those parts can last uh the life cycle. Mhm. So in order to prevent that, because they're not track parts, we are changing them with new ones to kind of have a, you know, a baseline as to what is to be expected from a new blade. And we're also looking for alternatives that uh are in the works to have them single grain crystal so that they don't um fall victim to to creep like they are currently. Um and also the HPT blades we're looking into to make sure that there's no, you know, um yeah, creep uh crack propagation that can happen. So far we we didn't see anything, but those are like the the two main components. Everything else that normally would fail under low cycle fatigue is kind of behaving the way we expect it and we don't necessarily foresee a uh Mhm. Mhm. a creep failure if that makes sense.

Mhm. Mhm. But but during testing, so which kind of uh problem has needed the most needed the most troubleshooting? Is there an example you can share of uh what went wrong, how it was fixed and uh and how the fix was checked? Yeah.

Yeah, no, that's a it's a fair question. Um the main issue that we had during testing was related to um the VBV system. Um essentially controls um obviously when you're testing different fuels, when you're testing different operating regimes and you are kind of redefining the operation um envelope of an engine or how fast you want it to accelerate or how slow you want it to accelerate or you know, there's there's a few parameters that uh come into play that you need to to kind of redesign or rethink about. Um and when we were testing uh the main issue that we had was in relation to the like the yeah, like I said, the the VBV system. Um because when you're normally starting up your engine with a different design in front, like the booster, um your inflow is different, right? Your inflow is different, therefore your N2 or your core reacts differently, therefore you have to adjust the fuel. Um but if you adjust the fuel in one way, let's say you put in too much fuel, then your core kind of stalls because your your VBV system isn't um operating the way it should or um because you're injecting too much fuel to kind of compensate for let's say a a fast start, um then your VBV system uh sends a signal to the ECU that tells it, hey, there's an issue with uh the core, so stop. And then you like it it's a weird uh PID system problem that you have to kind of figure out uh and reverse engineer. Um but that was definitely like the the main one. Uh everything else is kind of taken into consideration. So it's really yeah, VBV, VSV because your scheduling as well is a little different. Um but if you're able to control your your VBV, you're essentially able to um kind of nail down your your stop and start uh operation envelopes. So you can when it comes to after after that like max con, not much changes really. Um you're just monitoring your your N1 to make sure that you're not overspeeding and you're also just monitoring your N2 to make sure that you're not uh overheating.

Mhm. Uh okay. So so so for the VBV system, it is you you designed in-house or or it's from a supplier?

No, so the VBV system is still the the the OEM one. Like when we're talking strictly parts, um the the doors themselves are are all OEM doors. Uh what I'm talking about specifically is the um the control system around it. Um which is essentially just like logic gates, right? Um so the hardware is the same, but the logic gate behind it is slightly different due to a different intake. Right? And something I didn't really touch on as well is the, you know, like the the booster blades are one, but there's also like your inlet guide vanes that are slightly different that can also affect your flow, which is not something that you have on the OEM uh design. So that that's also why it really uh you know creates a challenge when you're trying to test it.

Yeah. But but I mean is it difficult to to to to find the best uh the the the VBV for the model one, right? Or or is it is it very difficult or after several testing then you can find the right the the the right way?

Yeah, exactly. Um it's it's it's trial and error. Um what I'm discussing about right now was over the span of like two and a half weeks of non-stop testing um where you're able to design or iterate over your your operation envelope because you have to take into consideration that like whenever an engine stalls, like there's there's a hefty uh list of procedures that you need to do to make sure that your engine is good to go again. Uh because that's not a normal um operation. So um obviously that takes time because every time you're you kind of overshoot or undershoot your your operating envelope, it it kind of stalls your your progress. Um but yeah, it's through testing, it's through iterations of okay, essentially this is my max fuel intake I can put uh at you know this acceleration rate and you kind of go through that transient phase.

Okay, I I see, I see it. Uh but uh this problem wouldn't affect the timetable, right? Your your first delivery in 4Q?

No, like what I'm talking about right now is something that we've done months ago. Um right now for this prototype at least, uh it shouldn't. Um like I said right now we're just fine-tuning or we need to fine-tune the um the operation envelope. So like we have like the the no-go zones if you will or like your your limit and then we're trying to iterate through if you think just of a cycle, this is the best way to think about it. So when you go from 0 RPM to let's say 4,700 RPM N1, that transient, that acceleration or that um that change in velocity or or RPM has specific uh parameters that are all interconnected that change throughout. And sometimes it's governed by your N1, sometimes it's governed by your N2. And that's the hard part to to kind of nail down and and create like a PID system around. So that's that's that was the hard part. Now we're there, so generally speaking it's it's a lot easier. We're just fine-tuning that that envelope to say essentially if we want to stay at a uh how can I say this? Like if you want to stay within your your efficiency um basis or like your guidelines I should I should say like you want to be injecting I don't know I'll say a number like uh not more than 2,000 pounds per hour of fuel uh to to raise it. Otherwise you tend to overshoot it or you tend to uh choke your your core or um starve it. So anyways, but yeah.

Mm-hmm. Mm-hmm. Okay. Yeah, yeah, that's very clear. And and the next question is about uh what emission control setup does the first commercial package need? Uh where does SCR fit? Is it ready for that package or is it part of a later upgrade?

Yeah, so like I mentioned, for the emissions, I can't really speak too much on that because I'm not uh as involved, I should say, uh when it comes to uh how you control those. Uh I know that's a big part of it is the combustion uh system. You can have dual injection, you can have water cooling, etc, etc, which are all on the table. But as it stands right now, the emissions as is are similar to like the best iteration that CFM has for a technical insertion engine. Um that's yeah, I I'm not too sure if

Yeah.

Yeah, I I can't really speak too much on that, so sorry.

Yeah, sure, sure. I I understand. So actually I I'm trying to understand so which uh parts or which certifications or which factors would hold up shipment or operation in the intended markets?

Which factors? Okay, yeah.

Are you aware which factors or which certification you need to get or approvals, yeah, yeah, any factors would hold up the shipment.

Yeah, I mean, the the main one is probably around emissions and loud uh like the the noise emissions. Like we're talking like um uh SFC and um and noise. Uh the noise reduction I think is pretty good. I don't think there's anything that we would be able to to get from a certification perspective. We're currently testing in a very, very barebone prototype um way in in Miami for the last I don't know how many months. Um and we're already kind of like um meeting the criteria for the noise emission of the city. Uh so if it's good for the city, I'm sure they would be good in industrial portion of um let's say the western side of the US. Um so noise emission I know we're good. SFC emission, like I said, I'm not super uh confident that I can speak on that just because I'm not super familiar with um the regulations and certifications from that side. But if it's

Mm-hmm.

Yeah, like like the the only thing I could say is if the emissions are good enough for the SCR when it comes to a technical insertion engine, the CFM56, then it's going to be to that level or lower. Um, yeah.

Mm. Okay, okay. It's clear. So and what what does a customer need to see before accepting a unit? Is a successful factory test enough or do they need a longer run on site?

I think if you if you want to be extremely uh conservative as a customer, you probably want to see both, right? Um the factory test we already kind of have. Uh longer run on site is going to happen soon from a power perspective specifically, the arrow is already been proven. Um so if I'm a customer before accepting a unit, yeah, I would for sure would like to have a factory test done and the the back end that says, hey, we were able to run it for at least a week and based on our calculations it could last you I'll say a number like a month without any stops. Um I think if I'm also a customer, I would like to know how easy it is to monitor and keep operating uh the engine. Uh but that's something we're also working on to to make sure that it's easy to to integrate.

Mm-hmm. Okay, yeah, yeah, yeah, I understand. And and um actually I I I want to understand the the the testing process from the customer side. So uh do do they just need to see the test report or they would send some people to your factory to to to witness those testing? And and for that's for the existing customer. And for the potential customer, um what would they do to to to to testify to test your your your your your modules, your your mod one uh engines?

Okay, well to answer the first part of your question, there's already a few customers that were obviously interested in the product that assisted some tests. Uh customers that were already confirmed, already put their name to uh um or I should say like money on the line to to have the the few uh few uh first units. Mhm. Um I I think it builds a lot of trust when you're able to show how strict and efficient the procedures are in in place to to make sure that you're able to deliver a uh minimum viable product within a certain timeline. Uh and it kind of gives um the the customers like confidence of you know within the the company of the uh the parameters that you control and how much you can essentially um make sure that it will get delivered. So from from that standpoint I think uh that's very important. Um and your second part of the question you were essentially what were you asking again for the second portion?

Uh the first question is like for the potential customers, uh uh what what would they do to to to uh to to how to say witness your testing or or to see whether your products is uh how to say is trustable?

Right, but I mean realistically speaking aside from being able to see it yourself and seeing the the process that goes behind it, it's only going to be the the um the execution and the results, right? Unfortunately the results you can't um not that you can't predict but you can't necessarily fall back on them when they're good until they are good, right? Uh and for a new product that emerges on the market, you won't be able to fully know to what extent um you know the product will will hit the niche um market that it needs until you know like it's going to have to be you know like a year or two worth of service, right? Um the way I the way I see the market standing right now is you have at least two years lead time to get any Siemens engine. Mhm. Um so you're now if you absolutely need the energy uh throughput at the moment you're stuck with two main options, right? The the old trusty reliable Siemens engine but you're kind of out of it for for two year uh two years. Or you can kind of go through that new route for mod one, get a unit within the next few months and you can worst comes to worst you run it and if you're not satisfied with whatever the throughput, uh support, whatever it may be, well at least you'd have something for the next year and a half let's say. Right? Until you can if you still desire go with Siemens. Um and best case scenario your mod one kind of blows it out of the water when it comes to price and efficiency and how modular it is and then you still have a winning product that you can still order. That's the way I see it.

Mm okay I see I see. Uh and yeah for for the customer side uh I think you have disclosed one major con- uh contract with uh Elon right? Elon Musk uh-huh. Uh and uh I'm not sure whether you can share that whether uh some of the other major CSP uh cloud service providers has seeing the test in your factory or or or or or not? Because I think most of them would find all the options for for for for for power engine supply, right? So including your Model 1. So so so I I I I'm I'm wondering whether those uh major CSP or AIDC developers have uh have have have have seen your products or whether have seen your your your test reports yet.

Um, I'm not uh aware of that at the moment. I don't really deal on the commercial side of things as much. Uh, I'm sure if you'd ask Samamud, he would probably tell you uh all about it. Um, but uh I mean I I work with him but I I don't necessarily um discuss commercial uh aspects of things as as uh as much as we talk about the technical the the technical side of things. Um, I'm sorry, I don't think I can answer that question. I'm sure you could probably check in the public reports what's uh what's been published.

Okay. Okay. Okay. Uh, yeah, I see. But from a logical side, it's like uh if you want a Siemens uh aero aero derivative, you need to wait two to three years. Then and the Model 1 is the only option left for for for for the customer, right? In the next one and a half year, right? Is this is this correct? Or is there any other aero derivative suppliers in the market?

There are some that are starting uh like you've you've quickly touched on like uh Elon Musk um saying that they were going to do some aero derivatives in house and whatnot. Um, I think it makes sense if you have the inventory and also like already a um, what's the word? In French it's roulement uh or essentially like your yeah it'd be like equivalent to throughput. Um, however, I just don't see how Elon just comes up with um thousands if not millions of parts of aero product to then convert within the next year and a half or two years. I think we were really well positioned in the sense that like uh at outside we already have products that allowed us to do conversions within the CM56 line and we kind of knew the limitations of the engine. Um and because we have what three, four testing facilities, we have a lot of um of testing data we can fall back on. So we kind of understand the engine. If you're starting from scratch, I think, and I'm I'm speaking for myself here, like the the limiting factor is not necessarily the hardware, it's everything that has to do with the control system. And I've been, you know, saying that for a while now, but uh if you're not familiar with how the engine behaves, it's it's really hard to just start from scratch and create a new one. Um, so I'm sure that Elon probably will make one, but uh I I don't know. I I can't really speak on that because I'm not involved, but I think he's going to realize it's a lot harder than you think.

Mhm. But but do you think like in in but but but Elon would produce the whole parts or they would just produce the the most constrained one like the blades?

You're talking about our our providers right now or?

Oh no, just like uh yeah, Elon says it's going to produce the the the the engines, right? But uh I I don't think uh he he's going he he's going to use uh he's going to producing the the engines uh all by himself, right? Maybe he just uh produce the most constrained one like the blades and uh use the uh for the other parts. It's it's all from the OEMs, right?

Sure. Uh, I just, again, for me, from an FAA perspective, it's kind of hard if you're not the type certificate holder to just create the most in-demand parts. Um, like let's say for example, some of the parts that are most in demand are life-limited parts, right? Or HPT blades and whatnot. And aside from CFM and Chromalloy, the CFM56, there's not that many people that can make them. Um, Chromalloy just started their HPT blade PMA parts. Um, and even they are not able to produce as much as they say they are. Um, and then CFM, the waitlist is, I mean, we still get like 30% of their kits every week, or every, sorry, every three weeks, four weeks, I can't remember. Um, but it's not something that you could just build from the ground up. And LLPs, let's say for example, the process, the manufacturing process is so rigorous that only the OEM makes them for a reason, because they have the the secret process and um, I guess like best practices behind it. And you can see what happens when you try to iterate. Let's say if you look at um, uh, is it Pratt? Pratt or Rolls that uh, I think it's Pratt that started to um, to do LLPs with uh powder metallurgy instead of just casting them. Um, and their LLP life is not um, up to the standard of what they expected. So the moment you change it, it it kind of becomes tricky because your manufacturing manufacturing processes come into play when it comes to the lifing of your parts. Um, so let's say he does the blades, great. I just I just don't know how that differentiates him from any PMA manufacturer that we currently work with, unless it's the super high-in-demand parts like HPT blades and LLPs, but I think it comes, I don't know, like if I if I had to guess, it'd probably come into the play of how do you, as a non-type certificate holder, um, validate that you're up to the um, the ICAs or like the continuous um, the instructions of continuous airworthiness, right? Um, because you have to prove that you're essentially the OEM to the FAA, which becomes a tricky topic.

Okay. I see. So, so in conclusion, it's it's it's going to be a big challenge for them to to to get it done in in within like two to three years. It's difficult, I think.

I I think so, because let's say like just the HPT blades, it took Chromalloy I think four or five years to get the approval from the FAA. Um, and like I said, if it's any other like airfoil, there's already a ton of providers that can do PMA parts, which we are already kind of dealing with and and talking to for power. We already have like, well, Pacific um, Pacific Aero is one of them where we we acquired because they have so much uh capability to to produce blades. So like the airfoil side, we're good. Um, but if you're talking to anybody else, I I don't see how this would just change the the whole game when it just becomes another PMA part sourcer, because the the ones in in contention are are the HPT blades and and your LLPs. And I don't see how the LLPs can just be manufactured by a third party all of a sudden.

Mhm. Okay. And and let's talk about the the supply chain of the of the of the new blades. Like you need to uh replace the older blades with the new blades for Model 1, right? And uh how many suppliers do you have for those blades? And uh are they ready to ramp up for your So for your for your 27 ramp up.

So just uh clarification here, we're not replacing all the blades with new ones.

Mhm.

Uh I guess you're talking about the fan blades?

Yeah, yeah, yeah, yeah.

Okay, so for the fan blades, yeah, it's the new new manufactured one and it falls back into the um the the providers that I was talking about at the beginning of the meeting which um have already kind of promised a a set throughput in I'll say mid November. Um which we already have um a few units for. So fan blades were good. Um booster blades were good. The only the other part that I did mention is the LPT stage two blades, right?

Mhm.

Those are new for the moment. Um but it's not going to be how can I say the OEM standard for too too long. We're going to go with an alternate route um which will be single crystal. Um and that's um that's that's in the play. I don't have like a set deadline for that. So for now LPT stage two blades new are not too much of a concern to get because you buy them directly from the OEM and they don't tend to be they're not as in demand as like other parts of the engine. Um and I sorry I forget the the last portion of your question.

Uh yeah, it's like uh I think like uh two to three months ago, people were worried about uh the supply chain of blades. They say the blades are going to be a uh a major constraint for for the conversion to model one, but uh for the time being you think you don't think it's a it's a major constraint, right?

Well, not really uh because the air foil side, there's nothing that stops us from using PMA parts. There's nothing that stops us from um doing our own in-house repairs. Um when you when you think about and I guess this comes from like an aero um scope, right? Like if you talk to I'll say like American Airlines, right? They tend to not put any PMA blades on their engine.

Mhm.

Whether it be compressor, whether it be LPT, whether it be HPT or even booster. Okay?

Mhm.

So for them, they are on the wait list of the OEM. They have to wait I'll say I don't know like a month to get three sets. So if you have to wait a month to get three sets, you can do at most three engines per month, right? So obviously when you think in that scope or in that optic, then yes, that does become a major problem. When it the way I see it is let's say you have the OEM you can rely on where you have like an MCPS deal with them as well that allows you to get a pretty good price on the OEM parts plus uh kind of priority on some of the the sets that they release. And you can also fall back onto your PMA providers, whether it be Pacific Aero or Chromalloy or Heico, whether you know, there's a bunch of providers that we deal with. So we can pull on each of them because we have the flexibility to use PMA parts where for Aero, if if you're dealing with a major airline that have set standards for their maintenance practices, it's it's kind of harder to get to, right? So that I I think the air foil side is pretty much locked in aside from like I said, your maybe like your uh your fan blades, but that's going to be like mid November. So we should be fine. And also the uh LPT stage two blades, which we will be replacing, but again, um right now they're not hard to um to get straight from the OEM as new. So because they're not that much in demand. the math. So I I from the aero perspective I don't think this is as bad as people would have the gut feeling to to have them be. I think what will be harder is going to be like the the parts you don't necessarily see. Um, like it's easy to to think about what's going to, you know, be the bottleneck effect for the stuff that you do see, but the things that you don't see are are the things that are harder to predict and kind of compensate for.

Mhm. Mhm. And uh okay, okay. Uh I think uh let's let's talk about the the the the the okay the the economics. So what what what what what what is the current purchase price of a used CFM56 that can be made into Model 1? And and what is the other approx- uh other the approximate cost of the other necessary parts you need to purchase to to to convert it into Model 1? Do do you have any idea?

Yeah, sure. I mean I don't have an idea of like the packaging, so I can speak um of the engine without necessarily the added package of the PTO shaft, generator and and gearbox, which I would assume is probably uh an added like $2 million. Um, but I I don't know really. Um, but for an engine, for a CFM56 engine, it all depends of the life it has and the condition of the core. This is mainly what drives your um your your value, right? Book value when we talk about it like that. Um, but generally speaking, let's say for a perfect candidate for uh for power, you have an engine that is probably $4 million upfront cost, uh a good um $3 million worth of of repairs and restoration in your core, um just because you replace uh most of the blades um that are already used and you restore them, so you do an overhaul on the core. Um, keep your LLPs and and whatnot, so you're at like like seven. Um, let's say another extra million for for units and accessories, QEC, you're at eight. Um, add in the upfront cost of your uh packaging, I said 2 million, I don't know if it's actually 2 million, but you're at like 10 million. Mhm. More or less. Um, and generally speaking a brand new engine or I should I shouldn't say brand new, an engine with high cycle counts, which is what makes it more valuable for aero, sells for around like 12 million if it's a good one. Like 10, 12. Um, but for for power, you're able to get an engine that doesn't necessarily need to have as many cycles um because you're more concerned about like the not the the low cycle fatigue but the creep, right? Um, so you're able to get an engine that's cheaper than uh like you were you would be able to get for for aero, and you're also able to sell it for a a profit margin that that tends to be higher. So mhm you're able to I guess like uh how can I say this, you take your upfront cost is probably the same as like a a full core restoration for aero, but then your profit margin is higher. So you're able to kind of dig down deeper in your your restoration that you want to put in to make sure that it lasts the life that you're selling it for.

Uh so by by using those PMA parts, uh so uh how how how uh what's the percentage of uh of purchasing price you would lower uh for for those PMA parts?

Yeah, that's a good question. Um, I don't think I can disclose it for like the main components, but I can say like for units, let's say, I'll say this because I was involved recently in like the creating a maintenance program for it, but like let's say your your VBV system, um, your ball screw actuator, when we're talking a OEM certified overhauled one versus a PMA uh overhauled one, the price difference is like let's say the OEM one is um $10,000, the PMA one would be like two to 3,000. Uh so the price difference is quite substantial. Um, if you do that with the parts that uh tend to be more expensive on the aero side, HPT blades, it's not as much the profit margin uh or sorry the um the cost margins aren't as uh crazy, but they are still very significant. Especially if a set costs you $2 million, you can cut down quite a bit. We're talking close to a million dollars less uh per set uh for HPT blades. Um obviously the LLPs is the thing that we don't deviate from. Um but yeah, if you can yeah, you can think of like HPT NGVs, same thing, very expensive as a whole set. Um individually I think they're the parts that are the most expensive in the engine.

Mhm.

You can get a good uh 40, 50% discount on them.

Okay, well that's that's uh that's a good price.

Yeah.

And yeah, and uh the next question is like what actually what is the actual unit output power of the model one? Uh uh the management has it's 25 megawatt, but when you put it into massive production, uh what is the actual uh unit power of of of of each unit?

It's close to 25, it's above 20 or or yeah.

So I mean for for the I don't think I can disclose everything uh just yet, but I can tell you for like the first prototype, uh we were able to reach 21 megawatts gross without any uh improvement to the combustion chamber nor the correct or I should say refined uh VBV system scheduling. Um, so I I I think the 25 megawatts is going to be expected and you have to consider as well that we're running in in Miami with probably the worst conditions you could ask for, which is high humidity and high heat. Um, so anything that is a better condition than this should perform better. Um, especially when it when your EGT is the one that drives your um your output. Like let's say for example, if we were testing in Montreal in January, the let's say it's minus 15 outside, your EGT was going to be around at max continuous is going to be around 800 degrees, but if it's during the summer, it's going to be like 850. So your EGT margin is quite it's correlated essentially close to one to one. Um, so the moment you go into a drier or not as hot environment, I I think it's fair to expect uh a lot more power from the engine because you can run it faster or harder for the same EGT.

So what's what's what's the target? It's so so what is the target output? 22, 23?

Well target output is 25 megawatts at least. Um but now like so if you're asking me, I I think 25 megawatts is is very fair. It's more going to be considered uh like the the thing that's going to dictate how good of a product it is is going to be around yeah emissions and how much fuel you actually need to put in, how efficient the engine is to generate those 25 megawatts.

Mhm. And and and about the test, you have you have run the complete set test or it's just combustion chamber test?

What do you mean by complete set test?

Uh it's it's a It's like a complete unit testing or you're just testing the engine?

Oh yeah, no, a complete unit testing with the whole engine, PTO shaft, gearbox, generator.

Okay, okay. That's good.

Yeah.

Okay. And just a moment. I still have several questions.

No, go for it.

Okay. And one moment. And um yes, I still want about the the the the the risks. So if the first delivery or the production ramp slipped by six months, it's it's an assumption. What would be the most likely reason?

Um okay, what would be the most the most likely reason? I think the most likely reason, that's a good question. Um I'm thinking worst case scenario, not necessarily the most likely, but um it would be some type of creep failure that we didn't foresee. I think it's unlikely. Um but I think this is what would essentially drive us down to like reiterate for for quite a bit. And I think that's the one of the only scenarios that would like make it slip by six months. Um six months is really big. It would be like a major redesign. Um especially at this stage we're at. So I I think it would have to be some creep failure that we didn't foresee. Um the most likely um scenario would be a design constraint when it comes to emission for the combustion chamber, I would say. Um I think that's the most likely one. I don't think it would be slipping by six months, however, because like I said, we have like different iterations of the Mod 1. I I think the one that um people think about is probably going to be the equivalent of what I think of like the the second uh iteration because the first one is not necessarily the prototype, but it's the the proof of concept and the proof that the aero derivative works and is able to achieve 25 megawatts. And then the second and third are essentially going to make everything more efficient and, you know, less emission. Um so I think the design of the combustion chamber is probably the the likeliest one to make it slip. Um but I don't think it's going to slip the whole production by six months. It's going to be maybe slipping the third or second iteration by a few months.

Mhm. Okay, okay.

It would take more time.

Okay. Okay, and there's one last question. So um where does FTI's work end and the Jera's begin? You you you have a JV with Jera, right? And uh and uh who owns the final system testing and the fixes?

Yeah, okay. So the FTI work ends I mean, okay, so right now the production or um build of the engine is in Montreal.

Mhm.

Right? So that's FTI. And then the engine gets tested and finished up with the PTO, uh generator and um uh gearboxes in Miami. And then all this gets packaged into what is in the headlines of, you know, like the the trailer with the the mobile um setup of the engine uh by Jera, right? Like the these are essentially like the people that do the final packaging. So the work ends at FTI when you have everything except the envelope or the shell around the engine, which is like the that trailer uh setup. So engine FTI work stop when the engine is fully built uh and when the engine is delivered to the Jera facility with the PTO. generator and uh uh the gearboxes to to package.

Mm-hmm.

You're asking me who owns the final system testing and yeah and what not. It it would be FTI as well. So the idea is you build it in Montreal, test it in Miami, ship it to Jera, they package it, and then we get it back to do the final testing.

Mm-hmm. Mm-hmm. And uh for for for Jera, do they have enough uh I would say uh so if you target 100 units, okay, uh in in 27, uh uh do you see any constraint in in in the ramp up from Jera's side? Do they have enough uh capacity to to to ramp up uh in in the assembly line? That's a good question. I'm not uh as uh involved as the uh as other people when it comes to the packaging. Um, but from what I've heard, um, everything seems to be going fine. Uh, I think the the main quirks are setting up the um uh the materials that's required to to drive that assembly line, but I think the capabilities of the facility is um is up to our standards uh as it stands. Obviously the the packaging didn't start yet. So obviously they have they still have time to uh to uh to adapt. Um, but from what I've heard everything seems to be pretty much as expected. They're pretty um, from what I heard as well, like they're pretty um uh efficient when it comes to to making changes to to any packaging actually. They're they're pretty up there. Mm-hmm. Mm-hmm. Okay. And uh there's one last question. So uh recently Crusoe walked away from Boom's super power order. So what what's your view on that? Uh FTI is also new to those uh uh to to stationary power. So which risks are similar for Model 1 and which are different? Uh, yeah, I think the main um, sorry, you said Crusoe, right? Yeah, yeah, Crusoe, yeah, yeah. Crusoe walked away from a a major Okay, yeah, yeah, okay, I see. Yeah, okay, you're talking about that. Yeah. Uh, yeah, that was pretty recent actually. But um, yeah, I I I'm not too sure why. Uh, I should have read on that. I I don't know why they uh they necessarily stepped away. Do you know why? I don't know. So I I'm trying to ask you, right? Uh, from your view what why they walked away and uh do are there any similar risks for Model 1? That's that's what I try to figure out. I mean, all I know is that they wanted something that wasn't necessarily just turbine or aero derivative is what I what I heard. Uh, why? I'm I don't know. But uh when it comes to I I don't know exactly like all the details of the program of the of the order contracts either, right? Like if you're if Crusoe expected them to get it by I'm going to say, I don't know, next year, mid next year and then Boom Supersonic said, well, the best we can do is early 2028. That may be a contributing factor, right? I think time is a generally speaking like deal velocity is real. Um, if your velocity of delivery is not to the expectations of the uh in this case like the customer, um, generally speaking they they tend to find quote unquote like a better option faster. Um, and that tends to drive their decisions. I I think from a aero derivative perspective, I don't I can't think of anyone that is able to ship as fast as we do right now. So I think that's probably our our main advantage. Um, the only reason why we'd see a customer walking away is not necessarily for a timeline perspective. It'd probably be uh for maybe like uh efficiency perspective or field support perspective or Maybe like flexibility perspective, like if somebody comes up with a new hydrogen way of, you know, generating energy that tends to be cheaper, like that's the only way I could really see that at the moment. Um,

But do you see it as a major, but do you see, uh, see, do you see, do you see it as a major risk, uh, maybe some new hydrogen player coming out?

Uh, I just said that because one of my good friends is, uh, is doing this. I don't think the energy density of hydrogen is to the same level, but I'm trying to think of like something that's green, something that can generate energy and that is readily available. Um, that's the first thing that came to my mind. I don't think it's on the same level. Um, I'd be thinking that it's probably like a mix of a few things. Like let's say you're a big, uh, contractor and you want your energy, maybe if your budget is high enough, you don't lock in fully on aero derivatives, maybe you go with a mix of, I don't know, um, like it depends on a lot of things, but you can do with instead of just locking in fully aero, you can do, um, solar, you can do battery, you can do a mix of things. You can even do like a mix of coal as well. Like I think at the end of the day, the people that want the, sorry, the energy delivery is not necessarily honed down to one specific way to get it. They just want something that gets it quickly, efficiently, to a good price, right? I don't think it really matters the end, like the end product. They just want something that's reliable and they can get it as soon as possible for the cheapest price, right?

Mhm. Okay. Uh, and I have one last question, okay? It's about the traditional aero, aero, AP, AP business. So, uh, talking about the PMA parts, so, uh, could you talk about the, the trend of customer acceptance of the PMA parts? Uh, because I think you have a, you have a, uh, pressure on your margin of your AP business, right? And I think, uh, by using PMA parts, you can, how to say, increase your margin, but, uh, um, is customer accepting it? And how is the, the original parts company, uh, how to say, uh, what are the original parts doing, I mean doing to, to, to, to push back the PMA parts?

Yeah, I mean, that's a very, uh, that's, that's a question that deserves a conversation on its own, but, um, to answer the first part, like, I think, um, if you talk about PMAs at the moment, it's probably, and I'm talking specifically for the CFM56 engine, uh, market right now, it's probably in the best place it's ever been, um, because the engine, um, it's kind of like towards the end of its life, right? Realistically, you have the 7B that's after, you have the LEAP, uh, you have a few engine designs that were released already that kind of are taking over the fleet. So when you're still trying to run those engines, you want cost-effective maintenance plans, you want cost-effective programs, you want cost-effective options. Um, and because the OEM is slowly discontinuing the production lines or the availability of the OEM parts, well, people are kind of forced to find other alternatives. And in the last year, I've seen a quite drastic shift from airlines, um, that used to, you know, never say the P word, the PMA word, um, and that would never, uh, even consider that as an option to strictly go for PMA engines. I can't disclose, but like let's say, I can tell you there's an airline that has been buying our engine that had a full PMA set within it. We're talking HPT NGVs, HPT blades, LPT stage 1 NGVs, um, what else? Uh, HPT shrouds. shrouds, um, HPT compressor blades, everything was PMA. They bought it, never advertised it, but everybody in the industry knows and everybody in the industry is, um, taking a look at the aircraft that is flying that engine to see, essentially they see it as a fleet leader, right? Um, and they're trying to, I guess, get a good sense of if this engine is going to be as reliable as any other engine in the fleet. And if the answer is yes, I'm pretty sure they're going to, um, be a bit more interested in that because the PMA parts, if you asked somebody five years ago, they would go, I would never put that in my engine because for whatever reason, it's not, um, it's not proven or whatever the argument may be. But right now I can see there's a lot of people that are not necessarily looking for the most expensive parts in PMA, but the things that are not necessarily as, um, important. Let's say for example, like a starter. EasyJet is on backfill for 50, if not 60 starters that they need for their engines and they're stuck. The supply chain is dead. Uh, or they're waiting on Honeywell, but they can't get anything. So they're forced to go to PMA. Like, it's, it's kind of like the name of the game right now. Um, so yeah, sorry, tangent, but PMA, it's probably the best it's been. And when it comes to the power alternative, the PMA stigma, I think, is only negative in the aerospace industry. Um, I don't think it's as, uh, how can I say, pejorative in the industrial industry, uh, or aeroderivative industry because it's not flying. And the main concern that operators or airlines have when it comes to flying with PMA parts is the safety portion of things, which by the way, I think is unfounded, but I beg to differ. Uh, so on the industrial aeroderivative side, I think it's a more appealing option. Uh, a lot more appealing.

Mm, okay, okay, that's very clear. Okay, I think that's all for today and I wish you have a nice day. Thank you. Bye.

Thank you. Have a good one.

Okay, bye-bye.

Bye.