一 / 項(xiàng)目背景與相關(guān)案例

I. Project Background and Case Studies

數(shù)據(jù)中心熱回收（Waste Heat Recovery）是將數(shù)據(jù)中心運(yùn)行過程中產(chǎn)生的廢熱進(jìn)行回收再利用的技術(shù)，可顯著提升能源效率并減少碳排放。以下是國內(nèi)外典型應(yīng)用案例和技術(shù)方向：

Data Center Waste Heat Recovery is a technology that involves collecting and reusing the waste heat generated during the operation of data centers. It can significantly improve energy efficiency and reduce carbon emissions. The following are typical application cases and technical directions both domestically and internationally:

在國外，北歐的一些國家對數(shù)據(jù)中心余熱進(jìn)行回收利用。2011年，芬蘭某公司研究通過不同的方式收集云計(jì)算數(shù)據(jù)中心產(chǎn)出的大量余熱，并將其用于附近區(qū)域供暖，形成了良好的廢熱利用閉環(huán)。在芬蘭和挪威等國家，部分地方政府與當(dāng)?shù)財(cái)?shù)據(jù)中心達(dá)成協(xié)議，將數(shù)據(jù)中心產(chǎn)生的熱量購買并用于周邊公寓的供暖。

Some Nordic countries recycle the waste heat generated by data centers. In 2011, a company in Finland explored various methods to capture the substantial waste heat produced by cloud computing data centers and utilized it for district heating in nearby areas, establishing an effective closed-loop system for waste heat recovery.In countries such as Finland and Norway, some local governments have reached agreements with data centers to purchase the heat generated by these facilities and supply it to surrounding apartments for heating purposes.

在國內(nèi)市場，北京某熱力集團(tuán)將數(shù)據(jù)中心余熱經(jīng)水源熱泵提升后，供給市政網(wǎng)直供及二次網(wǎng)供熱。結(jié)果表明，可以產(chǎn)生巨大的經(jīng)濟(jì)效益和環(huán)境效益；某園區(qū)建設(shè)發(fā)展公司將數(shù)據(jù)中心的空調(diào)冷凍水作為熱源,為附近辦公樓供暖，并通過供熱系統(tǒng)智能控制。

In the domestic market, a heating group in Beijing utilizes waste heat from a data center by elevating its temperature via water-source heat pumps, subsequently supplying it to the municipal primary network for direct supply and the secondary network for heating. The results indicate that this approach can yield significant economic and environmental benefits. Meanwhile, an industrial park development company uses the chilled water from a data center's air conditioning system as a heat source to provide heating for nearby office buildings, optimizing the process through an intelligent heating control system.

素材來源：制冷與空調(diào)，如涉侵權(quán)，聯(lián)系刪除

數(shù)據(jù)中心的總能耗中，制冷系統(tǒng)通常占 30%-40%，部分傳統(tǒng)風(fēng)冷數(shù)據(jù)中心甚至可能超過50%。

Within a data center's total energy consumption, cooling systems typically account for 30%-40%, with some traditional air-cooled data centers potentially exceeding 50%.

隨著AI計(jì)算需求的增長，高密度服務(wù)器（如NVIDIA Blackwell架構(gòu)GPU）的散熱需求進(jìn)一步增加，傳統(tǒng)風(fēng)冷系統(tǒng)的能耗問題更加突出。

With the growing demand for AI computing, thermal dissipation requirements for high-density servers(e.g., NVIDIA Blackwell architecture GPUs) continue to rise, making the energy inefficiency of traditional air-cooling systems increasingly pronounced.

二 / 項(xiàng)目概況

II. Project Overview

泰鉑科技深耕數(shù)據(jù)中心熱管理，特別是對現(xiàn)有數(shù)據(jù)中心節(jié)能改造有獨(dú)特的見解和強(qiáng)大的技術(shù)支撐。應(yīng)新加坡某App運(yùn)營商邀請，泰鉑科技因地制宜地為其定制了能源改造方案，在為業(yè)主節(jié)約電費(fèi)的同時(shí)降低了碳排放，同時(shí)收到了額外的收益。

Taybo has deep expertise in data center thermal management, particularly excelling in energy-efficient retrofits for existing facilities. Invited by a Singapore-based app operator, we delivered a customized energy retrofit solution tailored to local conditions. This initiative not only reduced electricity costs and carbon emissions for the client, but also generated additional revenue streams.

新加坡某租用數(shù)據(jù)中心的企業(yè)，總機(jī)架數(shù)量65個(gè)，目前使用的機(jī)架數(shù)量為37個(gè)，每個(gè)機(jī)架的功率為8kW。為降低租用數(shù)據(jù)中心成本，特邀請?zhí)┿K為其定制能源改造方案。經(jīng)過實(shí)地調(diào)研走訪，發(fā)現(xiàn)數(shù)據(jù)中心鄰近有一牛奶加工廠。該牛奶加工廠目前采用燃油鍋爐為其巴氏殺菌工藝提供65-75℃熱水，日用水量平均為78.4噸，每月的柴油費(fèi)用約為29000新元。

A company leasing data center space in Singapore has a total of 65 server racks, 37 of which are currently in use, each with a power capacity of 8kW. To reduce data center leasing costs, the company specifically engaged Taybo to develop a customized energy retrofit solution. On-site investigations revealed that a milk processing plant is located adjacent to the data center. This plant currently uses a fuel-oil boiler to supply 65-75°C hot water for its pasteurization process, with an average daily consumption of 78.4 tons, resulting in monthly diesel costs of approximately S$29,000.

三 / 方案研討階段

III. During the proposal discussion phase

四 / 節(jié)能效益測算

IV. Energy Savings Assessment

經(jīng)測算，實(shí)施改造后，數(shù)據(jù)中心的年節(jié)約用電量為10.7萬kWh，年節(jié)約費(fèi)用3.8萬新元（電費(fèi)單價(jià)0.355新元/kWh），減少碳排放43.5噸；牛奶加工廠年節(jié)約柴油139200L，增加熱泵與水泵耗電155448kWh，共計(jì)節(jié)約運(yùn)行費(fèi)用29.3萬新元，減少碳排放310噸。本項(xiàng)目共減碳量353.5噸，節(jié)約運(yùn)行費(fèi)用33.1萬新元。

After the implementation of the transformation, it is calculated that the annual electricity savings of the data center amount to 107,000 kWh, with an annual cost savings of S$38,000 (electricity cost per kWh is S$0.355), and a reduction in carbon emissions of 43.5 tons. The milk processing plant saves139,200Lof diesel annually, while increasing electricity consumption for heat pumps and water pumps by 155,448 kWh, resulting in a total operational cost savings of S$293,000 and a reduction in carbon emissions of 310 tons. The total carbon reduction of this project is353.5 tons, with operational cost savings of S$331,000.

附表：節(jié)能改造估算表

Appendix: Energy Retrofit Cost-Benefit Estimation Sheet

備注：出于對客戶信息保密，以上案例公開數(shù)據(jù)或有調(diào)整