Distributed Resilience: From Electricity Consumers to Stakeholders in Urban Energy Infrastructure

Edited by Martin Tzou, Chong-Yie Cheng

Earlier this year, Taiwan Power Company (Taipower) formally announced that it would no longer approve new power connection applications exceeding 5 MW for data centers in Northern Taiwan. While the directive is technical in nature, its implications are highly concrete: expansion plans by major operators such as Google, Microsoft, and NVIDIA have begun shifting southward, not due to land availability or market conditions, but because the northern grid can no longer accommodate additional electricity demand. The spatial footprint of Taiwan's AI industry is increasingly being reshaped by the constraints of its energy infrastructure.

For those familiar with the sector, this development is not surprising but rather the delayed surfacing of a long-standing structural bottleneck. While Taiwan's grid resilience enhancement initiatives are both necessary and well-intentioned, they operate on a decade-long planning horizon. In contrast, demand from artificial intelligence (AI) and high-performance computing is scaling on a quarterly basis. Cities cannot wait, and neither can the industry.

1. Structural Bottlenecks in a Centralized Power Grid

The core issue is not electricity generation. Taiwan is not experiencing a conventional power shortage. The challenge lies instead in the structural limitations of its centralized grid system, in which electricity cannot be delivered with the level of stability, density, and real-time responsiveness required by demand centers.

Extreme weather events, aging infrastructure, and geopolitical risks are simultaneously amplifying the system's underlying vulnerabilities. During the 2024 Hualien earthquake, substations in northern Taiwan tripped, temporarily disrupting power supply to the Hsinchu Science Park. During Typhoon Gaemi, large-scale outages in Kaohsiung lasted for more than 48 hours, forcing hospitals and medical facilities to rely on diesel generators for backup power.

While centralized backbone infrastructure must be strengthened, cities cannot afford to rely solely on it. The consequences of disruptions at the network's edge have been repeatedly demonstrated during power outages. True resilience lies in enabling end-user systems to operate independently when necessary. Microgrids, distributed energy storage, and localized dispatch capabilities can help sustain critical services, including healthcare, water supply, telecommunications, and essential public services, even when major infrastructure is compromised. These solutions should not be viewed as alternatives to centralized systems, but as indispensable complementary components of a resilient energy architecture.

At the same time, the traditional approach of optimizing solely for supply chain efficiency has accelerated the concentration of resources around AI-driven industries and semiconductor manufacturing clusters. While this model has enhanced industrial competitiveness, it has also contributed to growing regional disparities. As a result, there is an increasing call for greater regional self-sufficiency and emergency response capabilities. This trend further underscores the need to strengthen local distribution networks and decentralized infrastructure, ensuring that communities are better equipped to withstand disruptions and maintain operational continuity.

2. Virtual Power Plants: From Urban Resilience to Economic Governance

This same framework also offers a practical solution to two increasingly pressing challenges: the shortage of renewable electricity available to businesses and the rapidly growing power demand driven by AI development.

The concept of a Virtual Power Plants (VPP) is straightforward. It aggregates distributed energy resources scattered across a city—including rooftop solar installations on public buildings, campus energy storage systems, community EV charging infrastructure, and behind-the-meter battery storage—into a centrally coordinated and dispatchable energy network.

Under normal operating conditions, a VPP functions as a commercially viable "invisible power plant," matching local renewable energy supply with the growing demand from businesses. During periods of peak demand or grid stress, it serves as a stabilizing force that enhances regional power reliability and flexibility.

For local governments, the implications extend beyond energy management. Energy is no longer merely a resource allocated by central authorities; it becomes a strategic municipal asset—one that can strengthen urban governance, enhance resilience, and support investment attraction by providing businesses.

Kaohsiung City has already taken the first step by promoting the development of a VPP through a public-private partnership (PPP) model. This marks the first time a local government in Taiwan has sought to position energy coordination and dispatch capabilities as a form of public infrastructure for investment and development. While challenges remain, particularly with regard to commercial viability and integration with Taipower's grid system, the strategic direction has now been firmly established.

For the Hsinchu semiconductor cluster, the challenge is not simply securing additional electricity supply, but developing a city-scale approach to energy coordination. This includes regional energy storage, integrated cooling systems, demand response programs, and industrial-scale VPP capable of balancing increasingly complex energy needs.

Meanwhile, as the Central Taiwan Science Park and Southern Taiwan Science Park continue to absorb AI-related investments and industrial expansion, a critical lesson must not be overlooked. If high-density electricity loads are merely relocated without simultaneously building the regional energy management and dispatch capabilities needed to support them, the bottlenecks currently facing northern Taiwan are likely to re-emerge in central and southern Taiwan in the years ahead.

3. Leveraging Taiwan's Industrial Strengths to Build Viable VPP Business Models

Achieving this vision does not require Taiwan to start from scratch. In fact, the necessary hardware and technological capabilities are already largely in place.

Taiwan has developed a comprehensive and internationally competitive supply chain spanning battery energy storage systems (BESS), energy management systems (EMS), solar power installations, and EV charging infrastructure. One notable example is Gogoro, which has deployed thousands of battery-swapping stations across Taiwan. Each station effectively functions as a distributed energy storage node embedded within the daily life of the city—a deployment model that remains rare on a comparable scale globally.

At the same time, Taiwanese startups are already applying AI-driven energy management technologies to aggregate electricity demand from thousands of distributed sites into virtual demand pools, enabling real-time matching of renewable energy supply and consumption. This reflects the core logic of a VPP—and, importantly, it is no longer a theoretical concept but an operational reality.

The challenge is that these capabilities are currently deployed primarily within individual commercial applications and, in many cases, are being exported to overseas markets such as Silicon Valley and Japan. In other words, the energy integration capabilities that Taiwanese cities urgently need are increasingly being developed by Taiwanese companies for use elsewhere.

The key policy task, therefore, is to create market-oriented frameworks that can weave these distributed hardware and software assets into scalable and commercially viable VPP business models. Rather than imposing unnecessary regulatory barriers, policymakers should enable innovation and participation, allowing industry to address domestic energy challenges while simultaneously exporting integrated VPP solutions to global markets.

4. Institutional Bottlenecks: From Renewable Energy Policy to Urban Governance

Yet the real bottleneck has never been technology.

To this day, Taiwan continues to view distributed energy primarily as an extension of renewable energy policy rather than as a foundational component of urban resilience infrastructure. While the Energy Administration under the MOEA has recently launched incentive programs for community-based power generation projects, the policy framework remains largely focused on promoting renewable energy deployment. It has yet to address a more fundamental question: What role should distributed energy play in urban governance?

Should VPP be regarded as public infrastructure or commercial services? Should local governments remain merely electricity consumers, or should they become active participants in the operation and coordination of energy systems?

Until these questions are addressed, even the most advanced technologies and extensive hardware deployments will remain little more than disconnected pieces of a larger puzzle.

Conclusion

Taiwan's 2026 local elections will usher in a new cycle of urban governance. Expectations for the next generation of city leaders should therefore be clear: energy can no longer be treated solely as the responsibility of the central government.

VPP and microgrids should be incorporated into long-term urban infrastructure planning, while public-private partnership mechanisms can be leveraged to attract private-sector capital, innovation, and operational expertise. The objective is to make energy coordination and management a core municipal capability rather than an externally provided service.

This is not a distant vision. The technologies exist, the hardware is available, and successful precedents have already emerged. What remains is the willingness of local governments to embrace energy governance as part of their own mandate.

The pathway has already been laid out. The question now is whether local governments are prepared to take ownership of their energy future.