水泥脱碳路线图 · 2026 — 2050
Cement will not be fixed by changing fuel. Around 60% of its CO₂ leaves the kiln from the limestone itself — chemistry, not combustion. This page benchmarks the 2026–2035 and 2035–2050 pathways, globally and by region: where capital is finding a real return, who is buying first, and which parts of the plan are quietly behind.
水泥问题不能靠「换燃料」解决：约 60% 的二氧化碳来自石灰石分解本身，是化学反应，不是燃烧。本页对 2026—2035 与 2035—2050 两个阶段、全球与分区域的路线图做基准评估：资本在哪里拿到真实回报、谁在第一批买单、以及计划中哪些部分正在悄悄落后。
CaCO₃ → CaO + CO₂. Roughly half a tonne of CO₂ is released per tonne of clinker before any fuel burns. Concrete is the most used manufactured substance on Earth and has no substitute at gigatonne scale — Asia and Africa will keep building. So the sector cannot shrink its way out; it has to change its chemistry, its fuel and its specification at the same time.
碳酸钙受热分解，每吨熟料在燃料点火之前就已释放约半吨二氧化碳。混凝土是地球上使用量最大的人造材料，在「十亿吨级」上没有替代品，亚洲与非洲仍在大规模建设。行业不可能靠缩量脱身，只能同时改变化学、燃料与规范。
Not one technology — a stack, in cost order: kiln efficiency and digital control → alternative fuels → clinker substitution (LC3, calcined clay, ground glass, recycled fines) → new binders → carbon capture → recarbonation and design efficiency. The cheap layers carry 2026–2035; capture carries 2035–2050 and is the largest single lever by mid-century.
路线图不是一项技术，而是一个按成本排序的技术栈：窑炉能效与数字控制 → 替代燃料 → 减少熟料（LC3、煅烧黏土、磨细玻璃、再生微粉）→ 新型胶凝材料 → 碳捕集 → 再碳化与设计减量。便宜的层撑起 2026—2035；碳捕集撑起 2035—2050，并在本世纪中叶成为单项最大杠杆。
Three markets must be built at once: a supply market (calcined clay, SCMs, CO₂ storage), a demand market (procurement rules, offtakes, EPDs and a legal definition of “near-zero”), and a risk market (carbon price, contracts-for-difference, concessional capital). Technology is rarely the binding constraint. Bankability is.
必须同时建三个市场：供给市场（煅烧黏土、辅助胶凝材料、二氧化碳封存）、需求市场（政府采购、长期承购、环境产品声明与「近零」的法律定义）、风险市场（碳价、差价合约、优惠资本）。真正的瓶颈很少是技术，而是可融资性。
部署的十年：把已经存在的技术，装到全世界的窑上。
2026 是政策时钟真正开始走动的一年。
到 2035 年必须完成的事，以及各杠杆在这一阶段减排中的指示性占比。
减少熟料用量：LC3（石灰石煅烧黏土水泥）可在同等或更低成本下减排 30–40%，是最便宜、最大、最快的一块。
替代燃料：欧洲热替代率已超 50%，全球其他地区多在个位数，投资回收期通常 2–4 年。
能效、绿电与人工智能窑炉优化：投入小、回收快，通常两年内回本。
首批工业级碳捕集示范：数量有限，但决定 2035 年之后的可行性。
新型胶凝材料、再碳化与设计减量：份额尚小，但学习曲线在这十年成型。
Indicative lever mix, after GCCA & IEA pathways / 指示性配比，依据 GCCA 与 IEA 路径
Roughly four fifths of what must happen before 2035 uses technology that already exists and mostly pays back in under five years.
2035 年之前需要完成的事，约有五分之四依靠的是今天已经成熟、且多数在五年内回本的技术。真正稀缺的不是发明，而是执行与规范。
难啃的尾部：剩下的那一半，只能靠基础设施与新化学。
碳捕集从「示范」走向「舰队」
From a handful of plants to something on the order of 800–1,000 kilns. This is no longer a technology question but a pipeline-and-storage question: capture is only bankable where shared CO₂ transport and permanent storage exist.
从个位数的示范线，扩展到约 800–1000 条窑线。这已不是技术问题，而是管网与封存问题——只有在共享的二氧化碳输送与永久封存存在的地方，捕集才可融资。
Operating since 2025 · Brevik辅助胶凝材料的悬崖
Fly ash disappears with coal power; blast-furnace slag disappears as steel decarbonizes. The two materials the industry leans on today are being removed by other sectors’ success. Calcined clay is the only globally abundant answer — and it is a 2026 decision, not a 2040 one.
粉煤灰随燃煤电厂退役而消失，矿渣随钢铁脱碳而减少：今天最依赖的两种掺合料，正被其他行业的成功抽走。煅烧黏土是唯一全球性充裕的答案，而这是 2026 年就要做的决定，不是 2040 年。
Structural risk · 结构性风险新型胶凝材料切入主流
Electrochemical cement, calcium-silicate rock routes, carbonatable binders and geopolymers move from first plants to perhaps 10–20% of a regional market — mainly where clean power is cheap and specifications allow performance-based approval.
电化学水泥、钙硅酸盐岩路线、可碳化胶凝材料与地聚物，从首套装置走向区域市场的约 10–20%——主要落在绿电便宜、且规范允许「性能导向」认证的地区。
TRL 6–8 today再碳化，被正式计入
Concrete slowly reabsorbs CO₂ over its life and at end of life. Getting this measured and credited in national inventories is a stroke-of-the-pen tonne — cheap, real, and currently under-recognised.
混凝土在使用与拆除后会缓慢重新吸收二氧化碳。把它测准、并计入国家清单，是一笔「写进规则就存在」的减排量：便宜、真实，目前被严重低估。
Accounting lever用得更少，用得更好
Over-specification is endemic: structures routinely carry 20–40% more cement than performance requires. Design efficiency, longer service life and reuse are the cheapest tonnes in the whole roadmap — and the only ones that need no new plant.
过度设计普遍存在：结构中常有 20–40% 的水泥超出性能所需。设计优化、延长寿命与再利用，是整条路线图中最便宜的减排量，且不需要任何新工厂。
Demand side剩余排放
Even on an optimistic path, a residual of several hundred million tonnes a year remains around 2050. It must be matched with permanent storage or removals — which means the removals market and the cement market are financially joined, whether or not anyone plans for it.
即便按乐观路径，2050 年前后仍会剩下每年数亿吨的排放，必须依靠永久封存或碳移除来对冲。这意味着碳移除市场与水泥市场在财务上已经绑在一起。
Net-zero closureThe roadmap is global; the binding constraint is local. In one region it is the carbon price, in another the cost of capital, in a third the specification code. Read each region for what is actually stopping it.
约占全球产量一半
The largest single swing factor on Earth. Falling demand plus compliance carbon pricing does most of the near-term work; the open question is whether the carbon price ever rises enough to justify capture.
全球最大的变量。需求下行叠加强制碳市场，已完成大部分短期减排；悬而未决的是碳价能否高到足以支撑碳捕集，以及封存的规则与选址。
增长最快，熟料系数最低
Already the world’s best blended-cement performer and the natural home of LC3 — abundant clay, abundant demand. The gap is downstream: CO₂ transport and storage geology are largely unproven.
全球混合水泥做得最好的市场，也是 LC3 的天然主场：黏土充足、需求充足。差距在下游——二氧化碳输送与封存地质条件仍缺乏验证。
价格信号最强，首套装置最多
The world’s laboratory for first-of-a-kind capture — Brevik, Hanover, Obourg, Lumbres, Padeswood. The risk is not technology but cost pass-through, import leakage and the speed of CO₂ network build-out.
全球首套碳捕集项目的试验场。风险不在技术，而在成本能否向下游传导、进口是否泄漏，以及二氧化碳管网的建设速度。
采购驱动，政策波动
Home to most of the venture-stage binder companies and the biggest corporate offtakes. Federal demonstration funding has proven reversible — which makes private offtake and state procurement the durable signal.
新型胶凝材料创业公司与最大企业承购协议的所在地。联邦示范资金已被证明可能反复，因此真正持久的信号是私营部门承购与州级采购。
未来需求增量最大
The decisive question is cost of capital: the same project can carry two to three times the WACC of a European one, which alone can turn a viable plant into an unviable one. Africa must avoid locking in high-clinker kilns it will own for 40 years.
决定性变量是资金成本：同一个项目在这里的加权资本成本可能是欧洲的两三倍，仅此一项就能让项目从可行变为不可行。非洲必须避免锁定未来四十年的高熟料窑型。
替代燃料与煅烧黏土的先行者
Quietly among the most advanced regions on the cheap levers, with commercial calcined-clay lines already running. Under-covered in global reporting relative to what it has actually proven.
在「便宜杠杆」上其实走得很靠前，已有商业化煅烧黏土生产线运行。相对其实际验证成果，在全球报道中被明显低估。
Return in cement is not a single number — it is a shape. Some levers pay back from operating savings today; some only pay back if a carbon price or a premium buyer exists. The useful question is which of the two a given tonne depends on.
水泥行业的回报不是一个数字，而是一种「形状」：有些杠杆今天就靠运营节约回本，有些只有在碳价或溢价买家存在时才成立。真正有用的问题是：这吨减排究竟依赖哪一种。
煅烧黏土与 LC3 · 最高确定性
The rare lever that improves margin and emissions together, because clay replaces the most expensive and most emitting component. Return does not depend on a carbon price.
少见的「同时改善利润与排放」的杠杆：黏土替代的是最贵、也最排放的那一部分。其回报不依赖碳价。
Highest-confidence ROI替代燃料与协同处置
A waste-management business hiding inside a cement business. The constraint is rarely the kiln — it is local waste collection, sorting and permitting.
本质上是嵌在水泥业务里的固废处理业务。瓶颈很少在窑，而在本地垃圾收集、分选与许可。
Fast payback数字化与能效
AI kiln control, waste-heat recovery, grinding optimisation. Small percentages, but applied across every kiln on Earth and available immediately.
人工智能窑炉控制、余热发电、粉磨优化。单点比例不大，但可立即应用于全球每一条窑线。
Deploy now新型胶凝材料 · 风险与上行都在这里
Electrochemical, calcium-silicate-rock and carbonatable routes. The venture upside of the sector — and the place where an offtake contract, not a lab result, is the real proof point.
电化学、钙硅酸盐岩与可碳化路线。行业里最大的风险投资上行空间；真正的验证不是实验室数据，而是一份长期承购合同。
Venture stage窑炉碳捕集 · 最大杠杆，最贵的钱
The largest single lever by 2050 and the one with the least natural return. It becomes investable only when three things are contracted together: capture, transport, storage.
到 2050 年最大的单项杠杆，也是自然回报最弱的一项。只有当捕集、输送、封存三者被同时锁定为合同，它才具备可投资性。
Infrastructure-grade二氧化碳矿化与再生骨料
Injecting CO₂ into fresh concrete, or mineralising it into demolition fines, saves cement per cubic metre. Modest per unit, but it is retrofit economics — the fastest thing to install.
向新拌混凝土注入二氧化碳，或将其矿化到拆除废料微粉中，可减少每立方米的水泥用量。单位效果有限，但属于改造型经济性，安装最快。
Retrofit economicsCement is 8–13% of the cost of concrete, and concrete is 1–3% of the cost of a building. Doubling the price of green cement moves a project budget by well under one percent.
水泥约占混凝土成本的 8–13%，而混凝土约占建筑总成本的 1–3%。即便绿色水泥价格翻倍，整个项目预算的变动也远低于百分之一。——这是整份路线图里最被低估的一句话：绿色溢价的问题从来不是买不起，而是没人被要求买。
以商业回报做长期好事的钱。
Breakthrough Energy (Ventures & Catalyst) — the most consistent backer of first-of-a-kind industrial plants, explicitly structured to fund the unbankable middle.
Just Climate (Generation IM) — asset-owner capital aimed at heavy industry; led Terra CO2’s Series B.
Lowercarbon Capital · Khosla Ventures — early conviction in electrochemical and mineral-route cement.
Amazon Climate Pledge Fund · Microsoft Climate Innovation Fund — rare investors who are also the customer.
Cemex Ventures · Holcim MAQER — corporate venture arms; often the fastest route from pilot to specification.
Public de-risking capital — EU Innovation Fund, US DOE demonstrations, Norway’s Longship. Without it, no first plant exists.
突破能源（风险基金与催化基金）是首套工业装置最稳定的支持者；Just Climate 代表资产所有人资本进入重工业；Lowercarbon 与 Khosla 押注电化学与矿物路线；亚马逊与微软的气候基金罕见地「既是投资人又是客户」；Cemex Ventures 与 Holcim MAQER 则是从试点走向规范最快的通道。公共资金（欧盟创新基金、美国能源部示范计划、挪威 Longship）是首套装置存在的前提。
真正把设备装上、把产品卖出的公司。
Heidelberg Materials — Brevik, Norway: the world’s first industrial-scale carbon capture on a cement kiln, roughly 400,000 t CO₂/yr, in operation since 2025, sold as a captured-carbon product line.
Holcim — the broadest low-carbon product range and the largest portfolio of capture projects in development.
Cemex — commercial calcined clay, a mature low-carbon brand and an active venture arm.
Vicat — the K6 full-chain capture project at Lumbres, France.
UltraTech · Dalmia Bharat · Shree — India’s low clinker factor leaders; Dalmia has committed to carbon negative by 2040.
Anhui Conch · CNBM — China’s scale players, with early capture demonstration units.
Sublime Systems · Brimstone · Fortera · Terra CO2 · Ecocem — the challenger chemistries now moving into first plants.
海德堡材料在挪威 Brevik 建成全球首个工业规模水泥碳捕集装置（约 40 万吨/年，2025 年投运）；Holcim 拥有最广的低碳产品线与最多在建捕集项目；Cemex 已实现商业化煅烧黏土；Vicat 推进法国 K6 全链条捕集；印度的 UltraTech、Dalmia、Shree 在熟料系数上全球领先；中国的海螺与中国建材在规模与早期捕集示范上占位；Sublime、Brimstone、Fortera、Terra CO2 与 Ecocem 则代表正在建首套工厂的挑战者化学路线。
需求端才是这条路线图真正的开关。
Microsoft — the largest single signal in the sector: a multi-year agreement to purchase up to 622,500 tonnes of Sublime’s low-carbon cement, tied to data-centre construction. One buyer converting a science project into a bankable revenue line.
Amazon, Google, Meta — data-centre concrete as the first at-scale premium market.
First Movers Coalition — members committing to at least 10% near-zero cement and concrete by 2030.
ConcreteZero — corporate buyers stepping to 50% low-emission concrete by 2030 and 100% net-zero by 2050.
Public procurement — Buy Clean rules and European green purchasing; governments buy roughly a third of the world’s concrete.
微软是行业内最大的单一需求信号：与 Sublime 签署多年协议，最多采购 62.25 万吨低碳水泥，与数据中心建设绑定——一个买家就把一个科研项目变成可融资的收入。亚马逊、谷歌、Meta 让数据中心混凝土成为第一个规模化溢价市场；First Movers Coalition 承诺 2030 年前至少 10% 使用近零水泥；ConcreteZero 设定 2030 年 50%、2050 年 100% 的阶梯目标。而政府采购约占全球混凝土用量的三分之一，是最被低估的杠杆。
Not five technologies — five conditions. Each one, if missing, stalls everything downstream of it.
这不是五项技术，而是五个前置条件。任何一项缺席，其下游的一切都会停在原地。
Make low-clinker the standard product, not the special order. Build the calcined clay supply chain now, before fly ash and slag disappear. This is the cheapest, largest and fastest gigatonne in the entire roadmap.
让「少熟料」成为默认产品，而不是特殊订单。在粉煤灰与矿渣消失之前，现在就把煅烧黏土供应链建起来。这是整条路线图里最便宜、最大、也最快的一块。
Long-dated offtakes and book-and-claim turn a green premium into a credit-worthy revenue line a lender can underwrite. Voluntary pledges finance nothing; a signed ten-year contract finances a plant.
长期承购与「账面认领」机制，把绿色溢价变成银行可以承销的、有信用的收入。自愿承诺融不到一分钱；一份十年期的签约合同能建成一座工厂。
Treat transport and storage as a regulated utility serving clusters of plants, not as a heroic project each kiln must solve alone. Capture economics live or die on this single structural choice.
把二氧化碳输送与封存当作服务于产业集群的公用事业来监管，而不是让每条窑线各自英勇解决。捕集的经济性，成败几乎全押在这一个结构性选择上。
Product standards, EPDs and verified MRV that let a performance-based low-carbon cement be legally specified. Without a definition, there is no market — only marketing.
需要有约束力的产品标准、环境产品声明与可核查的监测报告核查体系，使性能导向的低碳水泥能被合法写进设计规范。没有定义，就没有市场，只有营销。
Concessional capital, carbon contracts-for-difference, offtake and carbon price stacked so first-of-a-kind returns clear the hurdle rate — and, critically, so cost of capital falls in the regions where the tonnes actually are.
把优惠资本、碳差价合约、长期承购与碳价叠加起来，使首套装置的回报越过门槛收益率；更关键的是，让资金成本在真正产生排放的地区降下来。
判断任何一项水泥脱碳主张的三个问题。
If the answer is no, it is a demonstration, not a business — which is fine, but price it that way.
如果 2030 年之后离开补贴就不成立，那它是示范项目而不是生意——这没问题，但请按示范项目来估值。
Clay, storage, transport, grid connection. Announcements are free; contracts are the only evidence.
黏土、封存、输送、电网接入——是签了合同，还是只发了新闻稿？公告是免费的，合同才是证据。
If nobody has agreed to buy the output at a known price for a known period, the project has no floor.
如果没有人以确定价格、确定期限承诺购买产出，这个项目就没有底。
Everything above is meant to be checked, not believed. These are the primary sources — roadmaps, regulation, company disclosure and the demand-side coalitions. Follow the ones that matter to your decision and read them raw.
上面的一切都是用来被核查的，而不是被相信的。以下是一手来源：路线图、法规、企业披露与需求端联盟。挑与你的判断相关的，去读原文。
Figures are given as ranges and orders of magnitude drawn from the sources above; production, emissions and cost numbers move year to year and differ by methodology. Where a number is decision-critical, open the source and take it from the original table rather than from this page.
本页数据以区间与量级形式给出，来源见上；产量、排放与成本逐年变动且方法学不同。若某个数字对你的决策至关重要，请打开原始来源，直接从原表取数。