Leveraging our core product strengths,we offer three key solutions to global clients:First,Smart Grid Optimization Solutions,which utilize recloser controllers and distribution automation systems to enable rapid fault location and automatic restoration, enhancing power supply reliability. Second, Industrial/Residential Energy Storage Solutions, tailored to diverse load demands, helping clients achieve peak shaving,backup power assurance,and energy cost control.Third,EV Charging Network Solutions,providing AC/DC charger selection, network design, and O&M services, compatible with regional standards like North America's NACS and Europe's CCS2. These solutions have been successfully deployed in factories,office buildings,residential communities,and public transport,effectively addressing customer pain points in energy utilization,power stability,and charging convenience.

Green AI Data Centre Power Solution

RW Energy 08/12/2026

Multi-Scenario Solutions Based on Solid-State Transformers (SST)

1. Industry Background and Pain Points

1.1 Industry Background

In the AI large-model era, the power density of a single data-centre rack has jumped from the traditional 8 kW to 50–100 kW+, and total electrical load is moving from tens of megawatts to hundreds of megawatts. Power systems that are highly efficient, space-saving and quick to deploy have become the core competitive advantage of computing infrastructure.

Green AI Data Centre Power Solution

1.2 Four Pain Points of Traditional Data-Centre Power Supply

Pain point	Description	Physical root cause
Long power chain	10 kV → line-frequency transformer → 400 V → AC/DC → isolated DC/DC → 48 V	Five conversion stages, each losing 3%–5%
Efficiency bottleneck	Full-chain efficiency of only ~86%	Cumulative losses across multiple stages
Excessive footprint	Substation + UPS + distribution panels stacked	Line-frequency transformer plus multiple devices
Slow deployment	12+ months of on-site construction	Dispersed equipment, complex wiring

1.3 Traditional vs. SST Power Chain — the Physical Difference

Traditional chain (5 conversion stages)

Step	Equipment	Loss	In plain terms
①	Line-frequency transformer	1%–2%	Steps 10 kV down to 400 V
②	AC/DC rectifier	2%–3%	Converts AC to DC
③	Isolated DC/DC	2%–3%	Isolated DC conversion
④	Distribution busbar	0.5%	Power delivery
⑤	Board-level power	3%–5%	Steps down to CPU voltage

Total loss: ~14% across the chain, efficiency of only ~86%.

SST direct-connection chain (2 stages)

Step	Equipment	Loss	In plain terms
①	SST solid-state transformer	1%–2%	Directly connected at 10 kV; transformation, rectification and isolation in one step
②	Board-level power	3%–5%	Steps down to CPU voltage

Total loss: ~9% across the chain, efficiency of ≥90.8%.

Physical root cause: SST medium-voltage direct connection eliminates the line-frequency transformer and the intermediate conversion stages, removing the 3%–5% loss of each eliminated stage.

Green AI Data Centre Power Solution

2. Solution Overview and Value Proposition

2.1 Solution Definition (BLUF)

Replacing the traditional "isolated DC/DC + AC/DC + line-frequency transformer" chain with SST solid-state power delivers:

Full-chain efficiency of ≥90.8% (+5%)
Footprint reduced by 65%
Copper content reduced by 90%+
Installation time reduced by 75%

2.2 Core Value Proposition

Value dimension	Metric	vs. traditional
Full-chain efficiency	≥90.8%	Traditional ~86%, +5%
Footprint	−65%	One integrated module
Copper content	−90%+	Medium-voltage direct connection removes low-voltage high-current busbars
Installation time	−75%	Factory pre-fabrication plus plug-and-play
Reliability	N+X redundancy	Automatic bypass on single-module fault

3. System Architecture

3.1 System Topology (Layer by Layer)

① Medium-voltage connection: dual-loop 10 kV incoming — with incoming protection and metering
② SST power modules (N+X redundant): 10 kV AC → DC busbar → 400 V / 48 V output, with automatic fault bypass
③ Distribution layer: column header → racks — servers / storage / AI accelerators
④ Management layer: environment monitoring and energy management

Layer	Composition	Power flow	Design rationale
① Medium-voltage connection	Dual-loop 10 kV	Grid → SST	Redundant dual-supply feeding
② SST supply	N+X modules	10 kV → DC → Load	Integrated conversion and redundancy
③ Distribution layer	Column header + racks	DC/AC → Servers	Local distribution to cut losses
④ Management layer	Environment monitoring + EMS	Information flow	Energy-efficiency monitoring

3.2 Comparison with the Traditional Chain

Stage	Traditional	SST	Benefit
Voltage conversion	Line-frequency transformer + AC/DC + DC/DC	Single SST unit	Eliminates 3 conversion stages
Supply architecture	Dispersed stacking	Modular and integrated	Footprint −65%
Construction	On-site wiring	Factory pre-fabrication	Time −75%

4. Core Equipment and Technical Parameters

Parameter	Value	Physical significance
Full-chain efficiency	≥90.8%	Saves the loss of 3 conversion stages
Footprint	-65%	Integrated module
Copper content	-90%+	Medium-voltage direct connection
Installation time	-75%	Factory pre-fabrication
Reliability	N+X redundancy	Automatic fault bypass
Power class	Modular and scalable	Suited to AI clusters

5. Technical Highlights and Physical Principles

5.1 Highlight 1: A Slimmer Power Chain

Physical principle: every AC/DC or DC/DC conversion stage carries a 3%–5% loss. The SST merges transformation, rectification and isolation into a single stage, cutting the chain from five stages to two and eliminating those losses.

Quantified benefit: +5% efficiency and a significant drop in PUE.

Green AI Data Centre Power Solution

5.2 Highlight 2: Medium-Voltage Direct Connection

Physical principle: the traditional approach needs low-voltage high-current busbars (400 V, thousands of amps) to deliver power, with heavy copper loss. The SST connects directly at 10 kV, delivering power at high voltage and low current, so busbar copper usage collapses.

Quantified benefit: copper content −90%+, with lower heat generation and reduced fire risk.

5.3 Highlight 3: Plug-and-Play

Physical principle: modular factory pre-fabrication means only assembly and wiring are needed on site.

Quantified benefit: installation time −75%, letting AI computing centres come online quickly.

6. Economic Comparison with the Traditional Approach

Cost item	Traditional	SST	Savings
Civil works (footprint)	Baseline	−65%	Significant
Copper material	Baseline	−90%	Material costs slashed
Electricity (efficiency)	Baseline	+5%	Large annual savings
Construction	12 months	3 months	Early-launch revenue

7. Landmark Project References

Project	Application	Key data
GDS Chongqing	Direct-connected energy-storage integration	90% overall efficiency
CNPC Kunshan	Charging station	95%+ efficiency
Guangdong Power Grid	Flexible interconnection	98.7% efficiency

8. Technical Standards and Compliance

Standard	Scope
GB 50174	Data-centre design code
GB/T 34120	Energy storage connected to the grid
IEC 61850	Communication

9. Delivery and Services

Factory pre-fabricated modular units
Plug-and-play installation on site
Full-lifecycle remote operation and maintenance

Deployed Equipment

RWSST High Efficiency Solid State Transformer 8.5kV to 40.5kV

RWSST Series 840–3360 kVA, featuring ultra-high efficiency, outstanding safety & reliability, flexible combination and prefabricated design.The single system adopts N+3 redundancy to form an N+X architecture, delivering higher reliability. It optimizes power supply structure, achieves higher efficiency, faster deployment and more flexible power distribution solutions.Built-in prefabricated structure, cabinets are connected via busbars, which greatly reduces installation costs such as c

RW Energy
Modular Solid State Transformers for Adaptable Energy Integration

Product DescriptionSolid-state transformers (SSTs), also known as power electronic transformers, are based on the core concept of Power Electronic Building Blocks (PEBB). They generally employ a modular series-parallel topology to facilitate the flexible integration of varying voltage levels, capacity ratings, and forms of electrical energy, thereby enabling adaptable configurations for diverse application scenarios.FeaturesSmall size, few links, average efficiency improvement 3% (2%~5%)High fre

RW Energy
Modular Solid State Transformer for Data Centers based on PEBB Technology

TheData Center Solid State Transformer (SST), also known as the Power Electronic Transformer (PET), represents the next generation of green power distribution for mission-critical facilities. Built on the core philosophy ofPower Electronic Building Blocks (PEBB), it utilizes an advanced modular series-parallel topology. Unlike traditional transformers, the SST enables millisecond-level voltage regulation and fault isolation. It serves as an "Energy Router," providing flexible access for various

RW Energy