EnergyNet Explained: Internetification of Energy Distribution

Daniel Kammen; Franklin Carrero-Mart\'inez; Jimmy Chen; Joakim Wernberg; Jonas Birgersson; Marc A. Weiss; Michael Menser; Newsha K. Ajami; Tomas K{\aa}berger

REVIEW 4 major objections 5 minor 9 references

EnergyNet Explained: Internetification of Energy Distribution

T0 review · 4 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that electricity distribution can be rebuilt as an Internet-like network of software-routed DC domains, with galvanically separated Energy Routers and open protocols, and that this architecture is a feasible, superior repl

desk verdict A clearly-written architecture vision for DC microgrids, but its feasibility evidence is anecdotal and the firewall stability argument has an unresolved tension. read the letter →

arxiv 2509.08152 v1 pith:2ZGSWPMJ submitted 2025-09-09 eess.SY cs.SY

Jonas Birgersson , Marc A. Weiss , Jimmy Chen , Daniel Kammen , Tomas K{aa}berger , Franklin Carrero-Martínez , Joakim Wernberg , Michael Menser

show 1 more author

Newsha K. Ajami

This is my paper · ORCID

classification eess.SYcs.SY

keywords EnergyNetDCmicrogridsoftware-definedpowerroutinggalvanicseparationRouterProtocolnear-real-timedistributionInternet-inspiredgridarchitecture

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that the same architectural shift that turned circuit-switched telephony into the packet-switched Internet can now be applied to electricity distribution. It proposes EnergyNet: a software-defined, layered network of DC microgrids interconnected through power-electronic Energy Routers that enforce galvanic separation and negotiate power flows via an open Energy Protocol. The authors argue this blueprint is implementable today with existing converters and control software, would let local communities operate autonomously while remaining interoperable with the wider grid, and would deliver resilience, scalability, and predictable low costs. They present early municipal demonstrations in Sweden as evidence that the architecture works in practice. A sympathetic reader would care because the claim is concrete and testable: if correct, the grid's distribution layer stops being a one-way, brittle bottleneck and becomes a modular, digitally routed network of networks.

What carries the argument

The Energy Router with its galvanically separated, software-negotiated DC backplane is the load-bearing mechanism. Each port can be AC or DC, variable-voltage, and independently controlled; power crosses the router's boundary only when the two sides explicitly agree through the Energy Protocol, mirroring a firewall. This single device is claimed to simultaneously isolate local domains from grid disturbances, enable bidirectional and prioritized energy flows, allow port-by-port scaling, and host redundant Energy Supervisors running EROS and EP-Server. The open Energy Protocol is the negotiation language that makes the network of networks interoperable.

What would settle it

A live test of a small network of Energy Routers islanded from the AC grid, with no synchronous generator attached, that must ride through the sudden disconnection of its largest source and a bolted short-circuit on one port while keeping critical loads powered; if buffered DC routing cannot hold voltages and currents within limits in that test, the central feasibility claim fails.

Watch

Extended reading notes

Core claim

The central claim is that the architectural principles of the Internet—decentralized routing, open protocols, local autonomy, and a 'firewall' between domains—can and should be transferred to electricity distribution. The paper specifies the components that would carry this: the Energy Router, a rack-based power-electronics device with a DC backplane and variable-voltage ports that enforces galvanic separation; ELANs and EWANs, DC microgrids organized like LANs and WANs; the open Energy Protocol for interdomain negotiation; and a control plane of EROS, EP-Server, and ENMS for operator-scale management. The authors argue that digitally controlled galvanic separation solves frequency-stability

Load-bearing premise

Electricity can be routed, buffered, and firewalled like Internet data, with power electronics providing the isolation and stability services that synchronous machines provide today, at scale and within the operational rules of connected AC systems.

Editorial extensions

If this is right

Local DC microgrids could keep operating autonomously during upstream grid failures, providing critical loads with partial power instead of blackout.
New generation, storage, and EV chargers could be added port-by-port on demand, avoiding long interconnection queues and enabling pay-as-you-grow capacity.
Renewable-heavy distribution would no longer depend on synchronous inertia for stability, because galvanic separation and local buffering contain disturbances at domain borders.
Multiple competing EnergyNet Operators could share the same physical infrastructure, analogous to multiple ISPs, potentially lowering costs and increasing choice.
If the architecture scales, EV charging, data centers, and industrial electrification could be served by dedicated DC energy networks without waiting for legacy grid reinforcements.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A decisive test of the firewall premise would be a multi-router, islanded DC microgrid with no synchronous machine attached: if it cannot ride through a short-circuit or a large load step by software-controlled buffering and rerouting, the frequency-stability claim collapses.
The near-real-time buffering analogy has a limit the paper does not address: electricity buffer lifetimes are milliseconds to hours, not the arbitrary queuing delays of data packets, so 'packetized energy' works only if storage capacity scales with every negotiated exchange.
The architecture's strongest near-term niche may be behind-the-meter and community-scale DC networks that coexist with the AC grid, rather than full replacement; the paper's own examples are all additive parallel installations.
The open Energy Protocol would need a formal specification and conformance testing to deliver the claimed multi-vendor interoperability; the paper names the protocol but does not publish its message semantics.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short take: this is a proposal/blueprint, not a demonstrated result. The paper's contribution is a named end-to-end architecture for DC microgrids with open protocols and a network-of-networks operating model. That is a useful organizing frame, and the Tier-1/Tier-2 split is honest: they separate architecture from adoption-dependent outcomes. The writing is clear, the analogies are explained well, and the appendix engages seriously with the smart-grid literature. I'll give credit for that.

The soft spots are real, though. There is no quantitative evaluation, no protocol specification, no cost analysis. The pilots are author-adjacent (ViaEuropa is the operator) and report only press-release numbers like the Tamarinden 30% reduction, with no baselines or measurements. The strongest internal tension is in §§4.7–4.8: the Energy Router is said to be a galvanic firewall that makes local domains autonomous and 'never disturbed,' while also providing stabilizing services to the AC grid through its batteries and electronics. Those two claims pull against each other unless there is a control law or negotiation protocol that shows how the router can refuse or delay power transfers and still deliver frequency support. The paper gives no such detail, and 'near-real-time' is never quantified. That is the load-bearing weakness.

That said, this is an architecture paper. It is not pretending to be a lab study. For readers who want a well-structured description of what a DC microgrid with internet-style routing could look like, and who are comfortable treating the pilots as existence proofs rather than measurements, this has value. It would be a reasonable submission to a forum that publishes visions or design proposals—perhaps an IEEE magazine or a perspectives section—but not as a primary research contribution. The lack of any formal model or independent data means I would not cite it as evidence for feasibility, but I might cite it as an example of this design space.

My recommendation: if you review it, judge it on its own terms as a blueprint. The authors should be asked to either supply a protocol specification and control-theoretic analysis or reframe the paper as a policy/vision paper, not an engineering result. I'd send it out for review, mainly because the architecture is coherent enough to warrant expert reactions, but I would not accept it under a claim that it demonstrates feasibility.

Referee Report

4 major / 5 minor

Summary. The paper proposes EnergyNet, an Internet-inspired architecture for electricity distribution. It defines components including the Energy Router, ELAN/EWAN microgrid layers, an open Energy Protocol, EROS/EP-Server control software, ENMS operator software, and parallel 'Freedom Cables.' The claimed contribution is a coherent, open, and testable blueprint that replaces centralized AC distribution with software-defined DC routing, galvanic separation, local buffering, and negotiated interdomain exchange. The paper argues this yields local autonomy with global interoperability, near-real-time operation, resilience, and lower predictable costs. Feasibility is supported mainly by municipal demonstrators in Örebro and Lund, including the Tamarinden project and the SWS EnergyNet-0 pilot. The paper explicitly distinguishes Tier-1 architectural claims from Tier-2 adoption-dependent outcomes.

Significance. If the architecture worked as claimed, it would address real bottlenecks in distribution grids: interconnection queues, bidirectional flows, EV charging, and resilience. The Internet analogy is thought-provoking, and the modular, DC-centric, open-protocol direction aligns with ongoing microgrid research. However, the paper contains no mathematical model, no simulation, no measurement protocol, no protocol specification, and no independent data. The central feasibility claim is asserted rather than demonstrated. The authors are candid that this is a blueprint or how-to guide, but for a systems-engineering journal the absence of quantitative and falsifiable content is decisive. The paper could be a useful white paper or perspective, but it does not meet the evidentiary standard of a research article.

major comments (4)

[Section 4.7 (Firewall)] Section 4.7 states that the Energy Router enforces galvanic separation so that power flows only when both sides explicitly agree, making local resources 'never disturbed,' while simultaneously using the router's batteries and electronics to provide stabilizing services to the AC grid. These two properties are in direct tension. Frequency support requires sensing AC-side frequency and modulating real power in real time; that power order is a coupling path from the AC side into the local DC bus and battery. If the Energy Protocol may refuse or delay such transfers, the promised grid support is conditional and cannot be relied on by the AC operator. If it cannot refuse, local autonomy is not unconditional. The paper gives no control law, no protocol semantics, and no protection specification that would resolve this. This is a load-bearing contradiction for the feasibility claim.
[Section 4.8 (Near-real-time)] The paper argues that EnergyNet can operate in 'near-real-time' by buffering energy locally, analogous to Internet packet buffering. Unlike data, electrical loads require instantaneous power; local storage is finite, power-limited, and costly. The paper gives no definition of 'near-real-time' (milliseconds, seconds, or minutes), no storage sizing method, no latency bounds, no fairness or quality-of-service model, and no conservation constraints. The statement that packet-style buffering 'dramatically reduces complexity' is therefore unsupported. This is a central technical claim and cannot be accepted on analogy alone.
[Section 5.3.3 and 5.4.5] The feasibility evidence consists of the Tamarinden project reporting a 30% reduction in energy consumption and a 50% decrease in peak power demand, and the Lund SWS EnergyNet-0 pilot. The cited sources are a municipal project overview and a Viable Cities news post. The corresponding author is affiliated with ViaEuropa, described as the first EnergyNet Operator and the operator of the Lund pilot. No measurement methodology, baseline definition, weather normalization, load composition, cost data, or independent audit is provided. These anecdotal figures cannot validate the architecture, especially when the central firewall and buffering mechanisms are unspecified.
[Section 4.10 and 4.11] The paper claims an 'open, testable blueprint,' but the Energy Protocol is described only as similar to TCP/IP; no message types, negotiation states, failure modes, or security mechanisms beyond mutual TLS are specified. The Energy Router is described in terms of rack density and four port types, but there is no control law for voltage/current limits, no fault-current specification, no protection coordination with the existing AC grid, and no definition of the DC backplane's architecture. Without these details, the blueprint cannot be independently implemented or tested. This omission is load-bearing for the paper's stated contribution.

minor comments (5)

[Section 4.11.2] The claim that converter price-performance improves 'at rates comparable to the early stages of Moore's Law' is made without cited data or a defined metric. This is a strong quantitative claim and should be supported or softened.
[Figure 2 caption] The caption reads 'NASA satellite image of from November 24, 2022'; the word 'taken' or 'acquired' is missing.
[Section 4.7.1] The firewall analogy is overextended: an internet firewall controls information flow, whereas a power-electronic converter transfers energy and cannot inspect or reject a physical power imbalance. The paper should explicitly acknowledge this difference.
[References] References [18] and [20] are municipal and program websites, not measurement reports. Reference [12] points to a GitHub repository but does not state whether it contains a protocol specification or only project documentation.
[General] The paper uses many acronyms (POGS, ELAN, EWAN, EROS, EP, ENMS, ENO, BSS/OSS, eTOM). A consolidated nomenclature table would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: EnergyNet is an architectural blueprint with no derived predictions or fitted parameters to collapse into its inputs.

full rationale

This is an architecture/blueprint paper rather than a derivation. It contains no equations, no fitted parameters, and no quantity that is predicted from a model, so the standard circularity patterns (self-definitional, fitted-input-called-prediction, imported uniqueness theorem, ansatz-via-citation, renaming) cannot attach. The closest candidate is the reliance on the authors' own pilots in Section 5, especially Section 5.4.5, where 'the world’s first operational EnergyNet system' is described as developed 'by LKF together with the Lunds Kommuns Parkerings AB (LKP) Lund Parking Company and ViaEuropa' — ViaEuropa being the corresponding author's company. That is self-referential in an evidentiary sense and weakens the empirical support, but it is not circular in the logical sense required by the task: the pilot is not a fitted parameter renamed as a prediction, and the architecture's content is not defined in terms of the pilot. The Tier-1/Tier-2 distinction also hedges against falsification but does not make the architectural claims equivalent to their inputs. The firewall/near-real-time discussion in Sections 4.7–4.8 contains an unresolved engineering tension, but tension is a coherence and correctness risk, not circularity. No specific reduction can be exhibited, so the appropriate finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 6 assumptions · 5 invented entities

The paper's central claim rests on several domain assumptions about power electronics, regulation, and economics, and it introduces a new vocabulary for well-known DC microgrid concepts. No free parameters are fit. Independent evidence for the invented entities is absent: pilots are author-affiliated and the protocol is unspecified.

assumptions (6)

domain assumption Electricity distribution can be usefully modeled as a packet-switched network where energy is buffered and routed like data.
Enters in Sections 3.1 and 4.8; the entire architecture depends on this analogy being physically valid.
domain assumption Galvanic separation through power electronics fully decouples local DC domains from AC grid frequency and fault dynamics.
Section 4.7; if this does not hold at scale, the firewall claim fails.
domain assumption Open protocols and market actors can fund and operate energy infrastructure without subsidies, following telecom deregulation.
Sections 1 and 4.12; a policy and economic assumption, not demonstrated for energy.
domain assumption EU and Swedish regulation permits Freedom Cables parallel to incumbent grid infrastructure.
Section 4.15 and reference [15]; a legal interpretation that underpins the deployment path.
domain assumption Power electronic converter costs will continue improving at Moore-like rates, making router ports cheap enough for redundancy.
Section 4.11.2; cost trajectory is extrapolated, not measured.
domain assumption Internet resilience after cable cuts transfers to energy networks.
Section 4.1; an analogy from RIPE measurements, not a physical equivalence.

invented entities (5)

Energy Router
purpose: Central power-electronic gateway with galvanic isolation, DC backplane, and variable-voltage ports.
Real hardware exists in the authors' pilot, but no independent third-party measured characterization is provided.
Energy Protocol (EP)
purpose: Open standard for negotiating energy exchanges between routers.
Protocol semantics are not specified and no implementation or commit is shipped.
ELAN and EWAN
purpose: Logical energy network hierarchy analogous to LAN and WAN.
Conceptual abstractions; no standard or interoperability test is provided.
Freedom Cables
purpose: Parallel DC cables deployed when incumbent grid owners do not adopt EnergyNet.
A new name for parallel DC feeders; legal viability is asserted but no deployment data is given.
EnergyNet Operator and ENMS
purpose: Operator role and management system for fleets of routers.
Organizational and business model, not a validated technical system.

how reviews work

0 comments

Cite this review

Pith. "Pith review of EnergyNet Explained: Internetification of Energy Distribution." pith.science (2026). https://pith.science/paper/2ZGSWPMJ

@misc{pith2026250908152,
  author       = {Pith},
  title        = {Pith review of: EnergyNet Explained: Internetification of Energy Distribution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2ZGSWPMJ}},
  note         = {Machine review of arXiv:2509.08152}
}

read the original abstract

In developing EnergyNet we have leveraged and are extending lessons from telecom's shift from a centralized, circuit-switched phone system to decentralized, packet-switched data networks. EnergyNet utilizes 1) an Energy Router that enforces galvanic separation and utilizes software-controlled energy flows over a DC backplane, 2) Energy Local and Wide Area Networks (ELAN/EWAN) based on DC microgrids that interconnect through an open Energy Protocol (EP), and 3) a control plane comprised of the Energy Router Operating System (EROS) and EP Server which is managed at operator scale through an Energy Network Management System (ENMS). We distinguish the architectural contribution (Tier-1 including components, interfaces, and operating model) from expected outcomes contingent on adoption (Tier-2). The latter includes local-first autonomy with global interoperability, near-real-time operation with local buffering, removal of EV-charging bottlenecks, freed grid capacity for data centers and industrial electrification, as well as a trend toward low, predictable, fixed-cost clean energy. Evidence from early municipal demonstrators illustrates feasibility and migration paths. The contribution is a coherent, open, and testable blueprint for software-defined, decentralized energy distribution, aligning power-systems engineering with networking principles and offering a practical route from legacy, synchronous grids to resilient, digitally routed energy distribution systems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

9 extracted references · 9 canonical work pages

[1]

EnergyNet is a modular, open architecture for energy distribution, with these key elements: • the Energy Router (galvanic separation, DC backplane, variable-voltage ports)

Executive Summary: A New Grid Architecture for the 21-century This paper introduces EnergyNet, an Internet-inspired architecture for electricity distribution, specifying its components, interfaces, and operating model. EnergyNet is a modular, open architecture for energy distribution, with these key elements: • the Energy Router (galvanic separation, DC b...

work page
[2]

network of networks

Introduction & Purpose Across the globe, the energy transition is facing a paradox. On one hand, renewable energy technologies like solar panels, batteries, and electric vehicles are becoming more affordable and available than ever before. On the other hand, our legacy energy distribution infrastructure – the traditional Alternating Current (AC) grid, or ...

work page 1997
[3]

some power

Why EnergyNet? For a century, electricity was generated at a small number of large power plants and pushed outward over one-way, centrally coordinated networks. That architecture -- radial feeders, synchronous coupling, and centralized protection -- was optimized for predictable, top-down power flows. The transition underway is the opposite. Generation is...

work page 2022
[4]

no- single-point-of-failure

EnergyNet: the Internetification of Energy Distribution Just like the telecom industry once relied on the POTS, the world’s energy infrastructure still remains dependent on the POGS. Like POTS, the traditional energy grid was groundbreaking at its inception but has become a rigid, outdated infrastructure incapable of adapting to today’s technological inno...

work page 2022
[5]

Allmän Energi

Real-World Examples: Key Organizations and Projects Across the European Union, cities such as Lund and Örebro in Sweden have begun deploying new DC microgrids, creating local demonstrations of EnergyNet’s full potential. These pilots validate both technical feasibility and economic benefits, proving that EnergyNet can be scaled quickly, cost-effectively, ...

work page
[6]

get richer by becoming greener

Conclusion: From Possibility to Deployment The traditional grid has brought electrification to billions and powered a century of growth, but it was never designed for decentralization, bidirectionality, or fully mobilizing the revolutionary capabilities of modern electronics and software. What the Internet did for communication, scaling up global access t...

work page
[7]

The Rise of the Stupid Network: Why the Intelligent Network Was Once a Good Idea but Isn’t Anymore,

References [1] D. S. Isenberg, “The Rise of the Stupid Network: Why the Intelligent Network Was Once a Good Idea but Isn’t Anymore,” Computer Telephony, vol. 5, no. 8, pp. 16–26, 1997. [2] L. Pappinen, “Bredbandsbolaget Gets Breakthrough Contract with HSB,” InternetNews, 31 Aug 1999. Available: https://www.internetnews.com/it-management/bredbandsbolaget-g...

work page 1997
[8]

more intelligence in the center

APPENDIX: What about Smart Grids? The smart grid era began with high hopes [21]: smart meters, sensors, and upgraded SCADA/ADMS would give operators real-time visibility, finer control, and fewer outages. Those programs delivered real value, better data, faster restoration, and volt/VAR optimization, but they largely left the core architecture unchanged: ...

work page

Show all 9 references

[14]

Review and Outlook on GaN and SiC Power Devices: Industrial State-of-the-Art, Applications, and Perspectives,

M. Buffolo, D. Favero, A. Marcuzzi, C. De Santi, G. Meneghesso, E. Zanoni, and M. Meneghini, “Review and Outlook on GaN and SiC Power Devices: Industrial State-of-the-Art, Applications, and Perspectives,” IEEE Transactions on Electron Devices, pp. 1-7, 2024. doi: 10.1109/TED.2...

2024

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [1]

EnergyNet is a modular, open architecture for energy distribution, with these key elements: • the Energy Router (galvanic separation, DC backplane, variable-voltage ports)

Executive Summary: A New Grid Architecture for the 21-century This paper introduces EnergyNet, an Internet-inspired architecture for electricity distribution, specifying its components, interfaces, and operating model. EnergyNet is a modular, open architecture for energy distribution, with these key elements: • the Energy Router (galvanic separation, DC b...

work page

[2] [2]

network of networks

Introduction & Purpose Across the globe, the energy transition is facing a paradox. On one hand, renewable energy technologies like solar panels, batteries, and electric vehicles are becoming more affordable and available than ever before. On the other hand, our legacy energy distribution infrastructure – the traditional Alternating Current (AC) grid, or ...

work page 1997

[3] [3]

some power

Why EnergyNet? For a century, electricity was generated at a small number of large power plants and pushed outward over one-way, centrally coordinated networks. That architecture -- radial feeders, synchronous coupling, and centralized protection -- was optimized for predictable, top-down power flows. The transition underway is the opposite. Generation is...

work page 2022

[4] [4]

no- single-point-of-failure

EnergyNet: the Internetification of Energy Distribution Just like the telecom industry once relied on the POTS, the world’s energy infrastructure still remains dependent on the POGS. Like POTS, the traditional energy grid was groundbreaking at its inception but has become a rigid, outdated infrastructure incapable of adapting to today’s technological inno...

work page 2022

[5] [5]

Allmän Energi

Real-World Examples: Key Organizations and Projects Across the European Union, cities such as Lund and Örebro in Sweden have begun deploying new DC microgrids, creating local demonstrations of EnergyNet’s full potential. These pilots validate both technical feasibility and economic benefits, proving that EnergyNet can be scaled quickly, cost-effectively, ...

work page

[6] [6]

get richer by becoming greener

Conclusion: From Possibility to Deployment The traditional grid has brought electrification to billions and powered a century of growth, but it was never designed for decentralization, bidirectionality, or fully mobilizing the revolutionary capabilities of modern electronics and software. What the Internet did for communication, scaling up global access t...

work page

[7] [7]

The Rise of the Stupid Network: Why the Intelligent Network Was Once a Good Idea but Isn’t Anymore,

References [1] D. S. Isenberg, “The Rise of the Stupid Network: Why the Intelligent Network Was Once a Good Idea but Isn’t Anymore,” Computer Telephony, vol. 5, no. 8, pp. 16–26, 1997. [2] L. Pappinen, “Bredbandsbolaget Gets Breakthrough Contract with HSB,” InternetNews, 31 Aug 1999. Available: https://www.internetnews.com/it-management/bredbandsbolaget-g...

work page 1997

[8] [8]

more intelligence in the center

APPENDIX: What about Smart Grids? The smart grid era began with high hopes [21]: smart meters, sensors, and upgraded SCADA/ADMS would give operators real-time visibility, finer control, and fewer outages. Those programs delivered real value, better data, faster restoration, and volt/VAR optimization, but they largely left the core architecture unchanged: ...

work page