GRIDSHIELD: Verifiable Grid Feasibility and Manipulation Resistance in Privacy-Preserving Peer-to-Peer Energy Trading

arXiv:2611.05816 · cs.CR · Submitted 2026-11-04 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Security Radio. Generated commentary on the latest security and cryptography papers.

Nadia: We're looking at the paper: "GRIDSHIELD: Verifiable Grid Feasibility and Manipulation Resistance in Privacy-Preserving Peer-to-Peer Energy Trading".

Elias: Here's the gist of the paper. Blockchain-based peer-to-peer (P2P) energy markets let neighbours trade directly; however, concentrate unchecked power in the clearing operator Page 1 GRIDSHIELD, a P2P energy-trading system that clears sealed bids with an order-independent batch program, removing the welfare a compromised operator can extract by reordering trades, and proves the result grid-feasible via an aggregated congestion-feasibility proof (ACFP) Page 1.

Nadia: The Problem Addressed The paper addresses the gap between network-aware clearing and privacy-preserving designs by introducing GRIDSHIELD, which clears sealed bids with an order-independent batch program, eliminating gains from transaction reordering Page 2.

Elias: Prior work either respects grid constraints on plaintext allocations or hides bids while dropping the network constraint or requiring a trusted setup Page 2.

Nadia: GRIDSHIELD removes the welfare a compromised operator can extract by reordering trades and proves that the resulting allocation satisfies grid constraints without revealing it Page 3.

Paper summary: Nadia: So, GRIDSHIELD isn't just about hiding things; it’s about making sure the outcome is physically possible while keeping things private >

Elias: They propose an aggregated congestion-feasibility proof, which they call ACFP, and they show that every single grid constraint—thermal line capacity limits, bus voltage magnitude bounds—can be reconstructed homomorphically from commitments to each bus >

Priya: That sounds impressive because it means the operator can prove feasibility by publishing something much smaller than all those individual constraints >

Nadia: They reduce all the feasibility checks down to a single aggregated range proof, and this proof's size grows logarithmically with the number of constraints, which is really efficient >

Elias: And they show that each peer can independently recompute their allocation from the published data and verify it against this single aggregate proof, endorsing the round only if its own check passes >

Conclusion: Nadia: So, looking at the full title "GRIDSHIELD: Verifiable Grid Feasibility and Manipulation Resistance in Privacy-Preserving Peer-to-Peer Energy Trading," it really captures the three main things they solved >

Elias: It’s about clearing a sealed bid market so there is no ordering advantage, proving the result respects every grid constraint without revealing it, and leaving an unfair clearing open to challenge afterwards >

Priya: What this means practically is that participants don't have to trust the operator completely because the system itself verifies physical limits are met >

Nadia: It’s about moving away from a setup where one central entity can manipulate things for their own benefit, and GRIDSHIELD claims it closes that gap using these new tools >

Abdullahil Kafi, Md Zahidul Islam, Alvi Ataur Khalil

Southern Illinois University Carbondale

cs.CR

Submitted: 2026-11-04

Updated: 2026-11-04

License: http://creativecommons.org/licenses/by-nc-nd/4.0/

Importance score: 85/100

The gist: Blockchain-based peer-to-peer (P2P) energy markets let neighbours trade directly; however, concentrate unchecked power in the clearing operator Page 1 GRIDSHIELD…

Key concepts

Aggregated Congestion-Feasibility Proof (ACFP)
This proof allows the system to check every physical grid constraint—like line capacity or voltage limits—at once using only a single, compact range proof. It achieves this by reconstructing each constraint from simple per-bus commitments, drastically reducing the size of the verification required for feasibility checks.
Congestion-Responsibility Fairness (CRF)
CRF is a new metric that measures fairness based on physical contribution to congestion. Unlike traditional measures, it assesses whether power curtailment reflects each party's actual physical impact on overloaded lines rather than just consumption or equal division.
Pedersen Commitments
These are cryptographic tools used to hide data while allowing mathematical verification. In GRIDSHIELD, commitments made per bus allow the system to prove that the hidden power allocations satisfy complex grid constraints without ever revealing what those allocations actually are.
Order-Independent Batch Program
This mechanism clears sealed bids in a way that removes any advantage an operator could gain by reordering trades. Because the clearing process is order-independent, a compromised operator cannot strategically manipulate the outcome to their benefit.

Summary

Blockchain-based peer-to-peer (P2P) energy markets let neighbours trade directly; however, concentrate unchecked power in the clearing operator Page 1 GRIDSHIELD, a P2P energy-trading system that clears sealed bids with an order-independent batch program, removing the welfare a compromised operator can extract by reordering trades, and proves the result grid-feasible via an aggregated congestion-feasibility proof (ACFP) Page 1. p. 0

Where the paper says it

We present GRIDSHIELD, a P2P energy-trading system that clears sealed bids with an order-independent batch program, removing the welfare a compromised operator can extract by reordering trades, and proves the result grid-feasible via an aggregated congestion-feasibility proof (ACFP). Exploiting the additive homomorphism of Pedersen commitments, ACFP reconstructs each grid constraint directly from per-bus commitments, reducing feeder-wide feasibility verification over any public grid constraints to a single aggregated range proof over the hidden allocation, with a reveal-and-recompute challenge that catches feasible yet unfair clearings.

The Problem Addressed

The paper addresses the gap between network-aware clearing and privacy-preserving designs by introducing GRIDSHIELD, which clears sealed bids with an order-independent batch program, eliminating gains from transaction reordering Page 2. Prior work either respects grid constraints on plaintext allocations or hides bids while dropping the network constraint or requiring a trusted setup Page 2. p. 0

Where the paper says it

edu Abstract—Blockchain-based peer-to-peer (P2P) energy markets let neighbours trade directly; however, concentrate unchecked power in the clearing operator. Most prior work addresses only part of this: network-aware clearing respects grid constraints yet operates on plaintext allocations, while privacy-preserving designs hide bids yet either drop the network constraint, enforce it only inside an unauditable computation, or recover it at the cost of a per-circuit trusted setup.

GRIDSHIELD removes the welfare a compromised operator can extract by reordering trades and proves that the resulting allocation satisfies grid constraints without revealing it Page 3. p. 0
Where the paper says it

We present GRIDSHIELD, a P2P energy-trading system that clears sealed bids with an order-independent batch program, removing the welfare a compromised operator can extract by reordering trades, and proves the result grid-feasible via an aggregated congestion-feasibility proof (ACFP). Exploiting the additive homomorphism of Pedersen commitments, ACFP reconstructs each grid constraint directly from per-bus commitments, reducing feeder-wide feasibility verification over any public grid constraints to a single aggregated range proof over the hidden allocation, with a reveal-and-recompute challenge that catches feasible yet unfair clearings.

Core Technical Contributions

The authors propose three main contributions to the field Page 3:

** We propose an aggregated congestion-feasibility proof and show that each radial-feeder grid constraint, thermal linecapacity limits and bus voltage-magnitude bounds alike, can be reconstructed homomorphically from per-bus Pedersen commitments, reducing all feasibility checks to a single aggregated range proof whose size grows logarithmically with the number of constraints Page 3. This proof relies on the additive homomorphism of Pedersen commitments to reconstruct each grid constraint directly from per-bus commitments, reducing feeder-wide feasibility verification over any public grid constraints to a single aggregated range proof over the hidden allocation with a reveal-and-recompute challenge that catches feasible yet unfair clearings Page 3. p. 0

Where the paper says it

We present GRIDSHIELD, a P2P energy-trading system that clears sealed bids with an order-independent batch program, removing the welfare a compromised operator can extract by reordering trades, and proves the result grid-feasible via an aggregated congestion-feasibility proof (ACFP). Exploiting the additive homomorphism of Pedersen commitments, ACFP reconstructs each grid constraint directly from per-bus commitments, reducing feeder-wide feasibility verification over any public grid constraints to a single aggregated range proof over the hidden allocation, with a reveal-and-recompute challenge that catches feasible yet unfair clearings.

**

** We propose congestion-responsibility fairness, a new metric to measure fairness grounded in grid physics, which measures whether curtailment reflects each party’s physical contribution to congested lines rather than consumption or equal division Page 3. p. 1

Where the paper says it

Our contributions are three-fold: • We propose an aggregated congestion-feasibility proof and show that each radial-feeder grid constraint, thermal linecapacity limits and bus voltage-magnitude bounds alike, can be reconstructed homomorphically from per-bus Pedersen commitments, reducing all feasibility checks to a single aggregated range proof whose size grows logarithmically with the number of constraints. • We propose congestion-responsibility fairness, a new metric to measure fairness grounded in grid physics, which measures whether curtailment reflects each party’s physical contribution to congested lines, rather than consumption or equal division.

This metric is defined as CRF = 1 − (X Σ i γi − ρi) ∈ [0, 1], which captures disparities missed by conventional measures such as the Gini coefficient Page 3. p. 1
Where the paper says it

• We propose congestion-responsibility fairness, a new metric to measure fairness grounded in grid physics, which measures whether curtailment reflects each party’s physical contribution to congested lines, rather than consumption or equal division. It captures disparities missed by conventional measures such as the Gini coefficient.

**

** We reproduce three published privacy-preserving and gridfeasible trading designs and compare them with GRIDSHIELD on common market rounds, real distribution feeders, and a 240-node feeder Page 3. p. 1

Where the paper says it

It captures disparities missed by conventional measures such as the Gini coefficient. • We reproduce three published privacy-preserving and gridfeasible trading designs and compare them with GRIDSHIELD on common market rounds, real distribution feeders, and a 240-node feeder.

**

Mechanism of Feasibility Proof

The aggregated congestion-feasibility proof (ACFP) exploits the additive homomorphism of Pedersen commitments to reconstruct each grid constraint directly from per-bus commitments Page 3. This allows the operator to publish one commitment per bus, Cj = Com(pj, ωj) Eq. p. 3

Where the paper says it

(2), so a commitment to every constraint’s value can be built directly from commitments to the bus injections. The operator publishes one commitment per bus, Cj = Com(pj , ωj ) Eq.

(1), and then compute a commitment to constraint k’s value (Fp)k using the homomorphism Page 3. p. 3
Where the paper says it

(1), together with the settled trades. Since every coefficient fkj is public, raising each Cj to fkj and multiplying gives, by the homomorphism, a commitment to constraint k’s value (Fp)k: Yn j=1 C fkj j = Com (Fp)k, Ωk , Ωk = P j fkj ωj , (6) therefore, each verifier computes Eq.

The operator then publishes an aggregated Bulletproofs range proof π, showing all 2K margins across all K constraints are non-negative at once Page 4. p. 3
Where the paper says it

(7b) Lk commits to the “above the floor” margin and Uk to the “below the ceiling” margin; proving both of them non-negative is the pair of checks that certifies bk ≤ (Fp)k ≤ bk. The operator produces one aggregated Bulletproofs range proof [20] π, showing all 2K margins across all K constraints are non-negative at once, made non-interactive via Fiat-Shamir with a shared bit-width β, and posts ({Cj}, π).

Each peer independently recomputes Lk and Uk from the published Cj and public F, verifies π against them, and endorses the round only if its own check passes Page 4. p. 3
Where the paper says it

The operator produces one aggregated Bulletproofs range proof [20] π, showing all 2K margins across all K constraints are non-negative at once, made non-interactive via Fiat-Shamir with a shared bit-width β, and posts ({Cj}, π). Each peer independently recomputes {Lk, Uk} from the published {Cj} and the public F, verifies π against them, and endorses the round only if its own check passes.

Evaluation Results

The evaluation assessed GRIDSHIELD across four axes: ACFP’s proof cost and how it scales with feedersize, the fairness of curtailment, a head-to-head comparison with existing designs, and cost on a live permissioned ledger Page 4.

** Proof size and scalability showed that aggregating the feeder’s 239 line constraints into a single range argument cuts proof size from 8,365 KB to 35 KB, a ≈238× reduction Page 4. p. 4

Where the paper says it

GRIDSHIELD’s batch clearing (Section III-C) drives this to exactly zero, however, leaves room for infeasible distribution. Proof size and scalability Aggregating the feeder’s 239 line constraints into a single range argument cuts proof size from 8,365 KB to 35 KB, a ≈238× reduction (Figure 3a), and because the aggregated argument is logarithmic in the constraint count, the reduction only widens as more constraints are stacked.

The proof size scales sub-linearly with the number of lines because the aggregated argument is logarithmic in the constraint count Page 4. p. 4
Where the paper says it

A 256-bit elliptic curve would shrink this further, to roughly 8. Across feeders (Figure 3b), the range argument grows logarithmically with the number of lines while the per-bus commitments grow linearly, so overall proof size scales sub-linearly: an 8× larger constraint set (IEEE 123-node vs.

**

** Fairness evaluation demonstrated that under a clearing where a compromised operator favors a colluding ally, CRF Eq. p. 4

Where the paper says it

Moreover, since ACFP reconstructs flows homomorphically, the constraint matrices never become arithmetic-circuit constraints, therefore ACFP needs no circuitspecific trusted setup where a generic SNARK would. Fairness of curtailment Under a clearing where a compromised operator favors a colluding ally, CRF Eq.

(8) drops from 0. p. 4
Where the paper says it

Fairness of curtailment Under a clearing where a compromised operator favors a colluding ally, CRF Eq. (8) drops from 0.

17 to 0.09, a 46% relative fall (Figure 5), while the curtailment Gini coefficient barely moves Page 4. p. 4
Where the paper says it

(8) drops from 0. 09, a 46% relative fall (Figure 5), while the curtailment Gini coefficient barely moves, from 0.

**

** A comparison with existing approaches showed that GRIDSHIELD’s proof cannot exist above 100% by construction, whereas an MPC-based clearing accepts an allocation we constructed at 392% Page 4. p. 4

Where the paper says it

More important is what each verifier will accept: an MPC-based clearing [12], [13] that omits the network constraint loads a line to 155% of rating and still verifies; a sealed-bid ZKP design [11] that checks individual matches without checking checking aggregate volume accepts an allocation we constructed at 392%. GRIDSHIELD’s proof cannot exist above 100% by construction.

p. 4
Where the paper says it

More important is what each verifier will accept: an MPC-based clearing [12], [13] that omits the network constraint loads a line to 155% of rating and still verifies; a sealed-bid ZKP design [11] that checks individual matches without checking checking aggregate volume accepts an allocation we constructed at 392%. GRIDSHIELD’s proof cannot exist above 100% by construction.

GRIDSHIELD’s proof cannot exist above 100% by construction, whereas an MPC-based clearing accepts an allocation we constructed at 392% Page 4. **

Practical Deployment

The protocol was deployed on a Hyperledger Fabric v2.5.9 network with two organizations, two endorsing peers and a Raft ordering service Page 4. p. 4

Where the paper says it

We therefore deployed the chaincode on a Hyperledger Fabric v2. 9 network with two organisations, two endorsing peers and a Raft ordering service, and drove it with Hyperledger Caliper using eight

The practical capacity of this deployment is approximately 1,064 tx/s Page 4. p. 5
Where the paper says it

03s once the offered load is past saturation. Flat throughput together with rising latency is the signature of a saturated pipeline, therefore 1,064 tx/s is the practical capacity of this deployment.

A market round places 939 transactions on the ledger, which over the one-hour clearing interval is an offered rate of 0. p. 5
Where the paper says it

Against that capacity, the protocol’s own demands are modest. A market round places 194 transactions on the ledger, which over the one-hour clearing interval is an offered rate of 0.

054 tx/s, roughly twenty thousand times below saturation Page 4. p. 5
Where the paper says it

A market round places 194 transactions on the ledger, which over the one-hour clearing interval is an offered rate of 0. 054 tx/s, roughly twenty thousand times below saturation.

The verification cost is paid once per round, and the total size of information a verifier needs to hold for proof verification and proof verification is 38. p. 4
Where the paper says it

0 (a) MPC clearing sealed-bid ZKP GridShield 0 200 400 worst line loadin (% of limit) 155% 392% 100% (b) limit (100%) Fig. (a) Total size of information a verifier needs to hold for proof verification and (b) Accepted energy flow allocation.

1 KB Page 4.

Conclusion

GRIDSHIELD closes the gap between two lines of work that each conceded what the other provided Page 5. It clears a sealed-bid market so that no ordering advantage exists, proves the resulting allocation respects every grid constraint without revealing it, and leaves an unfair clearing open to challenge afterwards Page 5. p. 5

Where the paper says it

CONCLUSION GRIDSHIELD closes the gap between two lines of work that each conceded what the other provided. It clears a sealed-bid market so that no ordering advantage exists, proves the resulting allocation respects every grid constraint without revealing it, and leaves an unfair clearing open to challenge afterwards.

ACFP reconstructs each linear grid constraint homomorphically from per-bus commitments, reducing a whole feeder, over a set of such constraints, to one aggregated range proof with no trusted setup Page 5. p. 5
Where the paper says it

It clears a sealed-bid market so that no ordering advantage exists, proves the resulting allocation respects every grid constraint without revealing it, and leaves an unfair clearing open to challenge afterwards. ACFP reconstructs each linear grid constraint homomorphically from per-bus commitments, reducing a whole feeder, over a set of such constraints, to one aggregated range proof with no trusted setup, and CRF gives participants and regulators a way to audit curtailment against physical responsibility rather than consumption alone.

CRF gives participants and regulators a way to audit curtailment against physical responsibility rather than consumption alone Page 5. p. 5
Where the paper says it

It clears a sealed-bid market so that no ordering advantage exists, proves the resulting allocation respects every grid constraint without revealing it, and leaves an unfair clearing open to challenge afterwards. ACFP reconstructs each linear grid constraint homomorphically from per-bus commitments, reducing a whole feeder, over a set of such constraints, to one aggregated range proof with no trusted setup, and CRF gives participants and regulators a way to audit curtailment against physical responsibility rather than consumption alone.

ACFP certifies feasibility sets expressible as public linear constraints; the thermal and voltage instances we evaluate use LinDistFlow, exact for radial feeders, so nonlinear limits (exact apparent-power ratings, losses, meshed AC) fall outside the present form Page 5. p. 5
Where the paper says it

Two limits bound these results. ACFP certifies feasibility sets expressible as public linear constraints; the thermal and voltage instances we evaluate use LinDistFlow, exact for radial feeders, so nonlinear limits (exact apparent-power ratings, losses, meshed AC) fall outside the present form.

The throughput figures reflect a two-organisation network on a single host, not a geographically distributed consortium Page 5. p. 5
Where the paper says it

ACFP certifies feasibility sets expressible as public linear constraints; the thermal and voltage instances we evaluate use LinDistFlow, exact for radial feeders, so nonlinear limits (exact apparent-power ratings, losses, meshed AC) fall outside the present form. The throughput figures reflect a two-organisation network on a single host, not a geographically distributed consortium.

References

The paper cites numerous prior works related to P2P energy trading, blockchain technology, and privacy-preserving protocols Page 4. Key references include the systematic review on blockchain technology in the energy sector Page 4, works on order-fairness consensus like Themis Page 4 and papers detailing zero-knowledge proofs for energy trading and secure multiparty computation. p. 5

Where the paper says it

McCallum, and A. Peacock, “Blockchain technology in the energy sector: A systematic review of challenges and opportunities,” Renewable and sustainable energy reviews, vol.

These references establish the existing landscape that GRIDSHIELD seeks to advance Page 4. p. 3
Where the paper says it

IEEE 13-BUS Topology IV. EVALUATION We evaluate GRIDSHIELD along four axes: ACFP’s proof cost and how it scales with feedersize, the fairness of curtailment, a head-to-head comparison with existing designs, and cost on a live permissioned ledger.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Replace operator-dependent clearing with an order-independent batch program to eliminate gains from transaction reordering, which is shown to be zero in GRIDSHIELD. This prevents a compromised operator from manipulating welfare by reordering trades for profit or curtailing a rival.

  2. Implement an aggregated congestion-feasibility proof (ACFP) that reconstructs each grid constraint directly from per-bus commitments, reducing feeder-wide feasibility verification to a single aggregated range proof with no trusted setup. This allows for efficient, verifiable proof of grid safety without revealing the allocation itself.

  3. Introduce congestion-responsibility fairness (CRF), a metric scoring curtailment against each party’s physical responsibility for congestion rather than consumption. This captures disparities missed by conventional measures such as the Gini coefficient by measuring whether curtailment reflects each party's physical contribution to congested lines.

  4. Utilize reveal-and-recompute challenges to allow participants to challenge a round, forcing the operator to reveal bids and commitment blinds, which can then be used to recover the optimum welfare W star and prove fraud if a sub-optimal clearing was used. This closes the gap where "W s < W star yields a fraud proof of the withheld welfare W star - Ws."

  5. Integrate ACFP into AI systems for real-time grid management, allowing autonomous agents to verify allocation feasibility against public linear constraints efficiently using only per-bus commitments and public data, which scales sub-linearly with constraint count.

Abstract

Blockchain-based peer-to-peer (P2P) energy markets let neighbors trade directly, but concentrate unchecked power in the clearing operator. Most prior work addresses only part of this: network-aware clearing respects grid constraints but operates on plaintext allocations, while privacy-preserving designs hide bids yet either drop the network constraint, enforce it only inside an unauditable computation, or recover it at the cost of a per-circuit trusted setup. Neither considers what happens if an adversary captures the operator's clearing privilege to reorder trades for profit or curtail a rival. We present GRIDSHIELD, a P2P energy-trading system that clears sealed bids with an order-independent batch program, removing the welfare a compromised operator can extract by reordering trades, and proves the result grid-feasible via an aggregated congestion-feasibility proof (ACFP). Exploiting the additive homomorphism of Pedersen commitments, ACFP reconstructs each grid constraint directly from per-bus commitments, reducing feeder-wide feasibility verification over any public grid constraints to a single aggregated range proof over the hidden allocation, with a reveal-and-recompute challenge that catches feasible yet unfair clearings. We introduce congestion-responsibility fairness (CRF), a metric scoring curtailment against each party's physical responsibility for congestion rather than consumption. We evaluate GRIDSHIELD against three prior P2P energy-trading designs across four feeders, including a 240-node Midwest feeder with real hourly loads, on a live Hyperledger Fabric network benchmarked with Hyperledger Caliper. GRIDSHIELD proves an entire round feasible with a few-kilobyte proof and no trusted setup, rejects infeasible allocations two of the three compared designs accept, and sustains about 1064 tx/s.

Sources

Related papers