Discussion Paper:
Grid Singularity’s Implementation of Symbiotic Energy Markets

Abstract

Grid Singularity’s mission is to build a sustainable, inclusive, and democratic energy market that facilitates ultimate degrees of freedom for the individual and the energy communities, allowing you to consume, trade, or share energy based on your preferences of an energy type, location source, price, or trading partner. You are energy.

Grid Singularity Exchange is built as an open source GPL v.3 codebase with an interactive interface to simulate and operate interconnected, grid-aware energy marketplaces. The future market design implementation, termed Symbiotic Energy Markets, combines multi-attribute double-sided auctions, graph representations of the energy grid and the decentralised computation and verification enabled by blockchain. In the envisaged consumer-centric market design, spot, futures, settlement, and balancing markets are intertwined through the use of time slots, allowing market-driven pricing and accurate accounting of delivered energy. Individuals can optimise for their own multi-attribute objectives (e.g. green energy source, reduced energy bill, preferred trading partner) through degrees of freedom in the multi-attribute double auction with dynamic pricing. The grid is represented as a weighted graph, offering grid operators efficient management tools and allowing energy communities to interconnect and trade. The complexity of the matching algorithm is decoupled from the verification of transactions, with matching facilitated by third party matching algorithm providers, termed “mycos”, enabling the exchange itself to operate on a blockchain. New energy market participants are rewarded for providing valuable services such as data and algorithms, while the roles of established market participants (aggregators, grid operators, utilities, and regulators) converge towards the individual and the community.

I. Introduction

Grid Singularity is implementing Symbiotic Energy Markets in stages, adding or adapting modules to its codebase to take advantage of advancements in blockchain technology, increasing regulatory interest to enable bottom-up markets and the proliferation of connected energy resources. We rely on open source innovation and ecosystem collaboration to develop an inclusive, efficient, accountable and secure energy exchange:

1. Inclusive

  • to all stakeholders, providing barrier-free access [1] for all market actors (established and new) including individuals and their communities, grid operators and utilities, trading algorithm providers, aggregators/energy management service providers, regulators, as well as diverse energy assets;
  • to any individual choice, allowing participants to trade and interact with energy based on their personal preferences [1] on a wide choice of attributes.

2. Efficient

3. Accountable

4. Scalable

  • technically, structured modularly and without matching and arbitration turning into a computational burden [8] for scenarios with many participants;
  • organisationally, allowing democratic deliberation and decisions on system upgrades;

5. Secure

Section II explores how defined market requirements translate to the following functional capabilities:

a. validated market participation and asset ownership,

b. degrees of freedom (choice of energy source, trading partner, etc.) with multi-attribute bids and offers,

c. matching and arbitration by third party algorithm providers that can be fully decentralised,

d. a time-slot mechanism that allows for spot, futures, settlement (post-delivery) and balancing trading,

e. an accurate grid topology representation that takes into account physical constraints and is scalable,

f. dynamic grid fees that allow grid operators to manage physical grid constraints,

g. a blockchain implementation for verification and storage of transactions and the grid topology in a deployed network, with decentralised blockchain architecture to allow deliberation and decisions on system changes,

h. privacy and security considerations for the protection of sensitive user data,

i. a data and algorithm marketplace to inform and optimise market participation, and

j. interoperable architecture to integrate existing and new business models, leading to a transformation of energy market roles.

Section III outlines the path forward to the deployment of the Symbiotic Energy Markets concept.

II. Grid Singularity Exchange Functionality

A. Validated Market Participation

To be inclusive, the energy market infrastructure should facilitate unhindered onboarding and interoperability of many different asset types and participants with various technology levels, which entails standardising the interfaces and registration process.

Present Grid Singularity canary test networks and proof-of-concept local market deployments are managed through the interaction of aggregators, who specify the attributes and placement of assets in the grid based on asset owner preferences, activating a local market operated by a centralised exchange. In the final implementation of the Grid Singularity Exchange, trading will be facilitated through a blockchain integration to ensure fully decentralised, verifiable, transparent transaction management (see for instance Michael J. Casey and Paul Vigna, The Truth Machine: The Blockchain and the Future of Everything [11]). This implementation will integrate Energy Web Switchboard, an open-source tool for decentralised identity and access management designed by the Energy Web Foundation, to verify users, assets, and application developers. Identity management systems allow verifiable claims to be presented to verified parties [12]. Participants are verified as individuals or groups in a decentralised registry. Each asset may have a decentralised identifier (DID) created by the manufacturer which is transferred to the owner upon purchase, allowing the participant to trade in the energy market and perform other operations with a verified energy asset. Utilities, grid operators, aggregators, and other players can also be validated through this mechanism. Notably, this is in stark contrast to current blockchain implementations of peer-to-peer energy trading where participants and the price are pre-determined and the exchange centralised and operated by closed source software.

B. Degrees of Freedom

In the Symbiotic Energy Markets concept, bids and offers for energy include a set of attributes (e.g. encrypted unique asset ID, asset location, type of energy produced e.g. solar energy, or membership in energy club) and diverse requirements that reflect trading preferences, such as energy quantity, price range, energy source, geographic distance, or preferred trading partner.

Multi-attribute bids and offers have been thoroughly discussed by researchers (including [1] and [13]) as mechanisms to capture participant trading preferences, and the technical feasibility of multi-attribute double auctions has been validated [14]. Multi-attribute bids are best matched by multi-criteria matching techniques, e.g. multi-criteria optimisation. Previous research has suggested trading strategies for multi-attribute double auction markets [15], identifying the need for algorithmic trading provision for self-interested market participants by third party trading agents due to the volume and complexity of decisions [2], [16]. [2] similarly attaches a set of requirements to specify energy, price, and time slot to orders.

Figure 1. Grid Singularity Exchange: data flows from energy assets over aggregators and trading agents via APIs for matching. Energy Web Switchboard is to be used as a decentralised asset registry.

In the Symbiotic Energy Market concept, intelligent agents managed by aggregators make algorithmic trading decisions on behalf of participants, translating energy asset information and prosumer trading preferences into a requirements function. Bids and offers are submitted through Grid Singularity’s existing Asset API, modified to include attributes and requirements for each submitted order. Each market then stores the list of attributed bids and offers in the exchange’s order book to be matched. Figure 1 above shows the data flow between the different actors and market exchange in a Symbiotic Energy Market.

Figure 2. An example of a bid and offer with attributes and requirements. The bid submits three sets of conditions. As the second condition is fulfilled by the offer, the two orders are successfully matched, in this case for 0.8 kWh of photovoltaic (PV) energy (at a price between 21 and 25 cents as determined by the matching algorithm). A verification function performs this check. The function accepts a bid / offer pair as input, returns a <True> (green check) if there is a valid match, and returns a <False> (red x) if requirements are not met. If the function returns <True>, a trade is created. In a near-term centralised implementation, this verification function is integrated into the exchange code. In a blockchain implementation, it is to be deployed as a module of the parachain’s protocol.

The structure of the requirements function, including the multiple sets of conditions and the specification of a `max_energy` enables some interesting properties;

  • Bids and offers can be matched one-to-many, allowing e.g. a single bid to fulfill multiple offers in the same operation, improving liquidity in the market and reducing the number of calls to the verification contract.
  • Defining `max energy` and multiple sets of conditions allows bids and offers to specify an energy mix, e.g. 30% PV, 30% from a preferred trading partner, and 40% grey energy.

If the pair were not a valid match for any set of conditions, both bid and offer would remain active in the market until a new valid match is found, or they are annulled by the owner of the bid or expire. If the energy requested in a bid or offer is only partially matched and cleared, the residual energy available in the order will be submitted as a new bid or offer in the market. Allowing partial matching increases the liquidity in the market.

The verification process has a computational cost, proportional with the complexity of the bid and offer’s requirements functions. This cost will be incurred by the aggregators in the form of fees, encouraging reduced complexity to maintain efficiency.

C. Matching and Arbitration

Multi-attribute auctions require more advanced clearing algorithms than single attribute auctions (e.g. stock market). Previous researchers, such as [2], propose a linear programming approach with clearing intervals to match multi-attribute price-energy-temporal requirements. However, in a blockchain implementation, the computational cost would make this approach unfeasible at scale. This has been shown to be solved by having (potentially untrusted) off-chain workers run matching algorithms requiring either the use of trusted execution environments (enclaves) [5] or an on-chain verification which determines if the matching algorithm was run correctly [2]. Both conclude, however, that multiple off-chain workers would be required to check the validity of the solution vs. an agreed upon matching algorithm and arbitration technique.

We similarly propose to decouple the matching process from the exchange, but remove the need for redundant off-chain workers, instead turning to third party matching algorithm providers (termed “mycos”), interacting with the Grid Singularity Exchange through a Matching API. The term myco is coined by drawing analogies to energy from mycorrhisal networks; underground networks of fungi that connect individual plants together to share nutrients, creating a symbiotic relationship within the forest ecosystem. In mycorrhisal networks, established and older trees act as hubs, continuously supplying carbon and nutrients to the fungal network. Younger trees gain access to excess supply, assisting their initial growth stages [17].

Mycos read active bids and offers through the Matching API, submitting matches for validation. Since the exchange’s verification function must only check the validity of each match rather than the result of a specific matching algorithm, there is no longer a need for an agreed matching algorithm, arbitration technique, or redundant off-chain worker computation. As long as the requirement functions of bid and offer are satisfied and grid fees paid, a transaction is created.

To specify the terms of a bid and offer match, common arbitration techniques include taking either the limit or the discriminant of price as the match point [5], [18], but any price within the valid range could be part of an optimal solution [14]. This allows the mycos to employ any existing price-based clearing algorithm (e.g. Pay-as-Bid, Pay-as-Clear, Pay-as-Offer) or create their own novel clearing algorithm (e.g. reinforcement learning techniques) for the multi-attribute double auction. The matching algorithm could introduce clearing periods instead of continuous matching to mimic established energy market clearing algorithms (e.g. Pay-as-Clear). Figure 3 below shows the way mycos interact with the local energy market.

Figure 3. Flow diagram of the role of mycos (third party matching algorithm providers) in the Grid Singularity Exchange.

If the verification function validates a myco’s proposed match, a transaction is created and the myco receives its reward. If the match is deemed invalid, the myco must pay the computational cost of verifying the transaction. The myco must therefore post a collateral that could be slashed if the verification function determines fraudulent or inefficient behaviour. The reward could include transaction fees or the minting of tokens, calculated as a function of the complexity of the match, the matched amount or another objective in line with the (likely multi-criteria) market objective.

In the first iteration of the Grid Singularity exchange, a single myco will be in charge of matching for isolated instances or clusters of local energy markets. Grid Singularity will offer a set of mycos that employ standard matching mechanisms (including Pay-as-Bid) modified to account for multi-attributed bids and offers. If a deployed community wants to employ a specific matching algorithm or partner with a specific matching provider, external mycos will be verified using a standardised registration procedure that includes posting an amount of tokens as collateral for potential invalid trades. In the long term, the registration process will likely be replaced by an on-chain consensus algorithm, possibly allowing many mycos to collaborate / compete to make efficient matches across interconnected energy communities, although unnecessary redundant computation should be avoided.

D. Time Slots — Spot, Futures, Settlement, and Balancing Markets

  • Spot and Futures Markets

Energy is traded in kiloWatt hours (kWh), a measure of power multiplied by time. Established energy spot markets typically include a Pay-as-Bid or Pay-as-Clear auction mechanism where energy is traded in 15, 30, or 60 minute time windows, such as the EPEX spot market [19]. The Symbiotic Energy Market concept includes the 15 minute interval as a constraint, although this could potentially be removed later for continuous bids and offers for energy over any interval. Many local energy market concepts trade energy 15 minutes ahead for immediate delivery, but the market structure must accommodate different time horizons to account for e.g. intra-day, day-ahead trading [16]. A futures market allows energy to be traded ahead of time, increasing the predictability of energy consumption for participants and grid operators. [2] enables bids and offers to be submitted for the immediate time slot or future slots by integrating a 15 minute window as a requirement for each submitted order, creating a futures market alongside the spot market. As this predictability is valuable to the management of the grid infrastructure, energy prices (and grid fees) may also be more favourable to participants in a futures market.

Figure 4. Spot, futures, and settlement markets allow trading centered around energy delivery.

Similarly, in the Symbiotic Energy Markets, bids and offers can be submitted with any 15 minute time slot (past, present, future) as part of their requirements function, as shown in figure 4. Mycos providing matching services would consider the selected time slot in their matching algorithm, matching only bids and offers valid for the same 15 minute slot (figure 5). Thus, the validity interval of a bid/offer can be set with arbitrary advance (i.e. hours or days ahead), creating a futures market alongside the spot market.

Figure 5. Bids and offers must select a time slot as a requirement, specifying during which time the energy requested or provided must be delivered.

It could also be possible to submit block bids or block offers, by bidding or offering a constant amount of energy for each 15 minute slot over a specified period of time (e.g. offer 1 kWh every 15 minutes from 13:00–17:15). Block bids are currently popular in day-ahead auctions, but require slightly more intricate matching algorithms to match and clear.

  • Settlement Market

Spot and futures trading requires predictions of energy use at both the asset and network level, which are often associated with high prediction errors [20], [21]. Thus, although a balancing mechanism should deal with the energy imbalances on the physical level, financial exchanges might still not be settled, causing a mismatch between physical energy delivery and market accounting. Some market approaches create orders and matches based on the last 15 minute slot’s energy use (post-delivery, read by smart meters) to solve this problem [5]. Others allow the grid operator to set prices [2] or enforce penalties [22] for any deviation. The Netherlands has implemented a final gate closure time one day after delivery, allowing ex-post trading for balance responsible parties to trade imbalances in real-time [23]. Symbiotic Energy Markets concept allows similar post-delivery trading by all market participants.

Deviations between energy physically produced/consumed and energy purchased in the spot or futures market can be traded post-delivery by submitting a time slot in the past as a requirement attached to a bid or offer, creating a settlement market (figure 6). Mycos can match these bids and offers with their typical matching algorithm. This allows local deviations to balance without penalty (e.g. a local solar panel that overproduces matches its deviation to a local load that over-consumed), with the remainder filled at a market rate (likely a premium) by individual assets, balancing groups, or grid operators that provided balancing services during delivery. A specific limit may be introduced to represent the time interval in which post-delivery trading can be conducted.

Figure 6. Energy deviations (difference between purchased and delivered energy) can be traded in the settlement market instead of paying a penalty to the grid operator, reducing inefficient accounting. Any remaining deviation can still be penalised by the grid operator.
  • Clearing Intervals

Eisele and others [2] specify a clearing time when trades must be finalised prior to delivery, and set a deadline on the submission, deletion, or alteration of bids and offers. In Symbiotic Energy Markets, the matching is decoupled from the posting of bids and offers, and trades are cleared continuously (in batches or blocks) as they are submitted for verification. This removes the need for clearing intervals. Bids and offers can be submitted, deleted, or altered at any time, as long as they are not actively being or have already been verified as a match. If regulation or a specific market design requires set clearing intervals, the myco could be scheduled to run the matching algorithm at encoded points in time.

  • Balancing Market

Local energy markets provide an incentive for participants with flexible assets such as energy storage to act as local balancing providers, balancing supply and demand during energy delivery in real-time when there are unexpected deviations between supply and demand, which could culminate in a blackout [24], [16]. A battery, for example, could be rewarded for reserving a portion of its energy for balancing energy to discharge if nearby solar panels are temporarily and unexpectedly under-producing. In today’s electricity markets, the provision of flexibility from distributed energy resources has been limited by existing market entry conditions, for example minimum bid sizes, contract durations, lead-time procurement, and pricing rules. Many of these are due to technical barriers, such as small bid sizes leading to a high number of bids that can overload the existing accounting system. These barriers can be overcome with the decentralised automated coordination of the bidding, matching, and delivery processes, fulfilled by the decentralised mechanisms proposed by Symbiotic Energy Markets.

In financial markets, put and call options give the purchaser the right, but not the obligation, to sell or buy some stock within an agreed period of time [25]. Symbiotic Energy Markets allow transmission and distribution system operators, balancing responsible parties (BRPs), or any other participant to purchase the promise of balancing services, by extending the concept of options to energy trading of demand-side or supply-side flexibility. Options can be created in the same manner that bids and offers are posted, matched by mycos, and validated with a verification function. Flexible assets that fail to deliver promised balancing services can be penalised in the settlement market, for example if an electric vehicle unexpectedly disconnects from the grid due to their owner’s personal mobility requirements. For such assets, a monetary collateral could be required upon submitting an offer in the balancing market.

The concept of options exists as orders for balancing capacity in the current European energy market design [26], albeit allowing only market participants with large amounts of power (minimum order size typically ~1 MW) or Virtual Power Plants (VPPs, aggregations of many smaller assets) to participate. The components of such an order for balancing capacity can be adapted for the decentralised use case proposed here, visualised in figure 7;

  • Option type: demand (consumption) or supply (production)
  • Power: minimum and maximum power required to be supplied by the balancing provider
  • Location: balancing group or physical location
  • Price / kWh: the agreed price of the delivered energy, and the premium for the option
  • Delivery: periods of time that specify the power delivery profile
  • Validity period: the period of time the option purchaser has to exercise the option
  • Mode of Activation: automatic or manually activated
Figure 7. Configuration requirements of a balancing market option for transmission system operators and other parties to procure balancing services. The option is matched by mycos and validated by the verification function of the Grid Singularity Exchange.

The exact components selected for use in each local energy market may vary based on geography (grid topology and weather conditions) and regulation. The components suggested here allow for more flexibility than the standard existing products for balancing capacity, intended to take full advantage of the versatile balancing controls of distributed energy resources. The specification of delivery parameters allows balancing options to be established for different time frames relevant to fulfilling the different types of balancing requirements and asset capabilities, from automatic activation up to 30 seconds (FCR) to RR (semi-automatic or manual activation triggered with >15 minutes delay).

The option could include a control requirement in the ‘mode of activation’, such that the asset automatically produces or consumes the agreed amount of energy, in line with standards such as IEEE 2030.5–2018 Standard for Smart Energy Profile Application Protocol and supported in the services of companies such as Kitu Systems. Assets could also delegate control to another entity, a function available in Energy Web Switchboard, which system operators might require for some trades. An option could be purchased in the futures market and exercised (fully or partially) at any point between purchase and the delivery of energy, or not exercised at all if not needed, incentivising flexibility provision by distributed energy resources. The financial incentive of the option will likely be trading at a premium to the market rate for the same amount of energy.

E. Grid Topology Representation

We consider grid topology because its disregard would result in a market platform without specific objectives and local trading would not be prioritised [16]. The power grid has been modelled using elements of graph theory and complex network theory [27], [28], resembling scale-free graphs, where a few nodes are highly-connected, constituting the hubs of large clustered groups of isolated nodes, resisting individual node failure or attack [29], i.e. forming independent, self-sufficient microgrids.

In a graph representation of the grid, substations are specified as nodes and electric lines are sketched as edges [27]. Symbiotic Energy Markets extend this concept, representing the grid topology as a graph with each node representing a market (e.g. substation, community, house). Each node could have a weight, allowing grid fees to be applied at the nodal level. The edges represent transmission lines, and could be weighted by a transmission capacity or a grid fee to trade energy across that line. This graph representation allows communities to benefit from trading with other nearby microgrids [22] while capturing the physical constraints of the grid network. The graph representation of the grid topology can be rendered on the exchange as an adjacency list [31] (figure 8).

Figure 8. Left: representation of the grid network as a graph. Nodes here represent transformers or interconnection points where markets are most likely to be placed (blue circles) and energy assets (purple diamonds). Weights can be assigned to nodes to reflect symmetrical nodal grid fees, or to edges to represent the cost to trade across a specific line. Right: adjacency list representation of the same graph used to store on-chain.

This grid representation would not be strictly hierarchical as in the current Grid Singularity platform, but would allow loops in the grid topology to be accurately represented. Individual assets or groups of assets could submit bids and offers tagged to their local market and mycos can match orders between markets. This removes the need for bids and offers to propagate to connected markets to be matched, which solves a scaling issue.

The myco responsible for matching can develop algorithms that crawl the network to match bids and offers across markets to find optimal matches [32]. The exchange’s verification function checks the proposed match and the associated grid fees. The most efficient matching algorithm would be incentivised to match energy locally and minimise grid fees, reducing computational cost and increasing the total amount of energy matched and overall system efficiency.

F. Dynamic Grid Fees

Grid Singularity Exchange aims to account digitally for the physical exchange of energy occurring in the grid, which is independent of individual trading preferences and has physical limitations. The grid’s frequency and voltage, network line losses, and other metrics must be balanced. Residential customers’ energy bills typically include the energy consumed and a grid fee to cover network operation [33], namely balancing the discrepancies between the accounted energy and the physical power flow. To address network constraints, [22] suggests that distribution operators could directly modify or cancel orders, while [2] applies physical safety constraints (e.g. consumption or production limits) to the clearing mechanism.

Grid fees paid by consumers have traditionally been static, dominated by flat tariffs designed to cover utilities’ long-term average costs [34]. Instead, dynamic grid fees are a flexible pricing mechanism that captures the varying true cost of supplying electricity throughout the day and would maximise social welfare while satisfying physical constraints of the power grid including power balances [35]. In this regard, [33] proposed a grid tariff for residential customers more reflective of grid cost based on power (kW) instead of energy (kWh). If the market slot is sufficiently short (e.g. 15 minutes) and power does not vary egregiously within that time period, the grid fee can be applied to the energy transaction as a cost in Euro per kWh (energy based pricing). Although peak power based pricing (Euro per kW) could moderate large peaks, once a user peaks in a time slot they would be incentivised to keep their peak high. Energy based pricing incentivises the user’s power to stay moderate over the entire time period. Long-term investments in grid infrastructure must also be priced into the dynamic grid fees.

In this market concept, the grid infrastructure is represented by a graph. A symmetric grid fee could be applied to each node allowing the grid operator to internalise network losses and congestion into the grid fee mechanism [37] (figure 9). If the grid fee is positive, the buyer pays the grid fee to the grid operator on top of the revenue that is paid to the seller. In the case of negative grid fees, the buyer pays for energy at a discount and the grid operator pays the discounted charge to the seller. These fees allow the grid operator to be paid depending on line use and incentivise traders to lower grid congestion which in effect lowers their balancing costs.

It could also be possible to weight the edges of the graph with grid fees applied to trades across that line, where the grid fee rate depends on the place of delivery [36]. The price of the grid fee can be decided by grid operators based on the status of the network, and be used to balance supply and demand and incentivise predictable behaviour that supports the healthy operation of the grid (e.g. lower grid fees in the futures market than spot market). There is also the opportunity for algorithmic pricing, such as automated market makers made popular in decentralised finance (DeFi), to help determine grid fee pricing based on market signals. Wang et al [38] showed that graphs can be time-dependent, with grid fees applied at nodes and/or edges to implement any dynamic pricing mechanism.

Figure 9. Time-dependency and space-dependency of nodal grid fees. The letters W, X, Y, Z represent grid fees for their respective nodes at two subsequent time steps.

Grid operators can algorithmically set grid fees dynamically for each market using Grid Singularity’s Grid Operator API. Futures market trading can be encouraged by setting grid fees as a function of time and the amount of energy scheduled, inducing more predictable energy usage behaviour. Grid operators can also adjust grid fees during an active market slot without affecting completed transactions (see figure 10).

Figure 10. Grid fees could be changed reactively to volume, e.g. an asymptote with volume and infinite grid fee, rewarding participants for trading when price is lower.

G. Decentralised Blockchain Architecture

Blockchain architectures allow market participants to post bids and offers and be matched on-chain with smart contract verifications [16], [18]. Grid Singularity’s future blockchain implementation will incorporate the identity authentication from Energy Web Switchboard (as described above) to verify participants and use the data for settlement, while its exchange is likely to be established as a parachain on a Substrate-based network (e.g. Polkadot). Notably, neither Polkadot or the Energy Web Chain rely on a proof-of-work mechanism of validation and therefore do not overuse electricity.

Parachains send a representation of the state of transactions to be validated by a relay chain that coordinates and interconnects many parachains, reinforcing the security of the network. A parachain has its own transition function between states, defined by the configuration of modules that make up the network’s protocol. Transactions and data crucial to defining the state of the network must be stored on-chain and available to use for the purpose of consensus. The state can occasionally be finalised by aggregating data and transactions; blocks created before finalisation do not need to be referenced again by validators in order to operate the chain (see section on availability in The Path of a Parachain Block), reducing the load to on-chain storage, particularly valuable for a high throughput of energy exchanges.

The parachain may store limited required information about the market configuration (e.g. market slot length, fee structure) as part of the on-chain storage enabled by Substrate. Grid operators submit changes to grid fees or the topology of the grid through the Grid Operator API, while market participants post pre-signed bids and offers for energy through the Asset API, tagged with their local market linked to the relevant grid topology, and conditioned upon set requirements (e.g. price, energy source). These grid management signals and active bids and offers will be broadcast to the network similar to how signed transactions are broadcast to the Ethereum, Polkadot, or Kusama network today, and aggregated into an order book and grid topology representation that will be maintained in an off-chain storage solution such as IPFS with authorised access.

Symbiotic Energy Markets will model mycos (third party matching algorithm providers) as off-chain workers which run matching algorithms off-chain, then submit bid / offers pairs to be matched through the newly developed Grid Singularity Matching API with authorised access. This is similar in concept to the proposals by [3] or [5], but without requiring a trusted execution environment (TEE) or pre-agreed matching algorithm. However, if a local energy community or a regulator prefers a specific matching algorithm, a myco could then optionally run the selected algorithm on SubstrateTEE. Matches are instead validated via an on-chain verification function module added to the parachain base protocol that checks if the bids and offers’ requirements (e.g. price, time slot, energy source) are met.

Mycos, working off-chain, are incentivised to efficiently match and optimise energy traded to maximise their reward, while any negative network effect or malicious activity is penalised through collateral slashing. The concept of off-chain matchers has for instance been implemented in Gnosis Protocol V2’s “Solvers” mechanism, implemented in Cowswap.

The verification function will be implemented as a module of the parachain’s protocol, taking a list of proposed bid and offer matches from the myco as input, and validating that the requirements of each bid and offer are satisfied. Valid matches are created as transactions in a block on the blockchain, while invalid matches are removed. The network validators (best available consensus mechanism to be determined at the time of implementation to maximise security and privacy) are responsible for running the module’s verification code and authorising blocks. Network costs are distributed according to network logic to mycos, aggregators, buyers, and sellers. As noted above, validators and mycos can be slashed for any potential malicious activity such as invalid matches or verifications. Financial transactions can also be integrated to automate payments.

Figure 11. Information flow between aggregator, Grid Singularity exchange and matching algorithm provider (to be sourced through Data and Algorithm Marketplace). IPFS or RDBMS will likely be used for off-chain storage of the order book. The matching algorithms will be run off-chain or in an instance of SubstraTEE. The Grid Singularity Exchange will operate as a pallet on a Substrate-based parachain or another network that provides the best security and privacy combined with execution efficiency.

In deployments of individual microgrids with limited assets, a single parachain may be sufficient for all network interactions. With a scaled network of many interconnecting microgrids and varied voltage levels, the number of transactions may require multiple parachains or parathreads with localised metrics that communicate and interact, possibly integrating Layer 2 solutions. Considering the fast pace of development of scaling solutions the final choice will be made based on the best available technology at the time of deployment.

Importantly, as the Grid Singularity Exchange moves to a fully decentralised mode of governance, it will also benefit from blockchain functionality to introduce system upgrades to further optimise local energy markets.

H. Privacy and Security

Individual energy data should remain anonymous to secure the system and its participants. [2], for instance, proposes how billing could be organised without prosumers disclosing anything other than the billed amount. Prosumers could use anonymous accounts when posting offers, but as [5] stressed they would still reveal their usage profile, which is personal data. [5] and other researchers have presented alternative concepts that provide confidentiality of personal data while still delivering transparency of the auction process, which Grid Singularity will consider and integrate in the exchange development.

I. Data and Algorithm Marketplace

The analysis of market and grid data can provide valuable input:

  • historical data and signals, energy usage predictions, and network analyses;
  • trading or matching algorithms for aggregators and mycos, grid management monitoring and grid fee strategies;
  • community or grid optimisations using historical data in simulations;
  • creative representations of energy interactions between participants used in social or artistic endeavours; and/or
  • processing of data for regulatory audit.

Creation of this marketplace will generate a business model for groups that can perform data analyses and provision data. It will also allow prosumers, energy communities, and researchers in Grid Singularity’s platform to purchase datasets and services to improve their simulation results from proof-of-concept to deployment. Grid Singularity is developing an energy-specific instance of an Ocean Protocol marketplace, partnering with innovative data technology companies such as Rebase to foster a collaborative ecosystem of data and algorithm providers.

J. Transformation of energy market roles

New or adapted business models will rapidly grow and mature with the introduction of local energy markets powered by exchanges like Grid Singularity, allowing diverse participants to be rewarded for value-adding contributions to the market:

  • Prosumers

Art. 15.2.b of the EU Clean Energy Package now entitles the sale of self-generated electricity [39]. However, in many cases today, excess energy input injected to the grid by prosumers’ assets either receives a feed-in tariff or no financial compensation at all. Local energy markets allow excess generation or stored energy to be sold to neighbours or other grid-connected parties at a market price. Pure consumers also benefit from reduced energy prices or socio-economic benefits, such as increased community self-sufficiency. Prosumers as owners of flexible assets can provide balancing services and may also be willing to sell their personal energy data for use in algorithm development or research.

  • Aggregators

Aggregators can charge prosumers based on the amount of energy transacted, a subscription fee including auxiliary services (such as providing an app to monitor use and exchange of energy use), and/or hardware upselling or other services. The exchange matching and verification fees as well as grid fees may be absorbed in the final charge. Aggregators would also provide the service of onboarding communities to exchanges, including the provision of smart meters, data agreements, and the registration of assets to the exchange.

  • Exchange Operators and Network Validators

In centralised operation, a small exchange fee for each transaction would be charged, paid to those responsible for operating the exchange software platform. In a decentralised implementation, a token-based business model could reward the validation of network transactions based on the configuration of network protocols. The tokenomics will be developed based on the advancements in blockchain infrastructure including governance options.

  • Mycos (third party matching algorithm providers)

Mycos will be rewarded for the matches they propose that are validated as transactions. A small matching fee will be added to each bid and offer, which can either be set by the network, or decided by each bid and ask, similar to gas fees in many blockchain networks. The myco would also post a collateral, in case a fee is slashed for invalid matches.

  • Algorithm and Data Providers

Algorithm and data providers may receive a subscription fee, a one-time fee, or a fee each time the data or algorithm is accessed, depending on the type of offering.

  • Grid Operators

Grid fees could be set by grid operators dynamically to better manage the grid. Local governments can also collect any relevant taxes. All payments could be automated via blockchain or other modalities.

  • Utilities

Utilities may lose part of revenue to microgrids, but can also offer balancing and auxiliary services. They can also benefit from the exchange efficiency, likely providing offers with large amounts of energy to act as a balancing body at the community level. They could also assume the aggregator role.

III. Path Forward

While this paper specifies the blueprint of the Symbiotic Energy Market, its implementation will be modular and progressive, accounting for regulation, community or grid operator standards and preferences in a four-stage development process:

  1. Decouple matching from transaction validation and enable a spectrum of energy choices (degrees of freedom)
  2. Integrate a futures and settlement market
  3. Integrate a balancing market including demand and supply options and direct energy asset management
  4. Enact organic, accurate grid topology and dynamic grid management mechanisms

The concept of Symbiotic Energy Markets is collaborative in nature, inviting discussion, as well as further research and development, particularly in the following areas:

  • Business models and reward functions for market participants and service providers,
  • Selection and long-term operation of mycos, especially in large multi-nodal networks,
  • Analyses of market efficiency through community and individual key performance indicators (KPIs), as discussed in [40],
  • Alignment and synchronicity of the modular components of the exchange, e.g. ability to annul active bids or offers,
  • Maximising privacy and security of the exchange,
  • Algorithm design for matching, verification, trading, and grid management,
  • Additional mechanisms for grid operators to reward the distributed optimisation of the physical grid.

Authored by the Grid Singularity team: Fatuma Mohamed Ali, Colin Andrews, Andrea Bertolini, Aeron Buchanan, Christopher Dietrich, Sarah Hambridge, Ewald Hesse, Ana Trbovich and Spyros Tzavikas. We thank seasoned energy and blockchain experts, Trent McConaghy and Tim Shittekatte, for their valuable comments.

References

  1. Mazzola, Luca, Alexander Denzler, and Ramon Christen. “Towards a Peer-to-Peer Energy Market: an Overview.” arXiv preprint arXiv:2003.07940 (2020).
  2. Scott, Eisele, et al. “Safe and Private Forward-Trading Platform for Transactive Microgrids.” 2019.
  3. Mengelkamp, Esther, et al. “Trading on local energy markets: A comparison of market designs and bidding strategies.” 2017 14th International Conference on the European Energy Market (EEM). IEEE, 2017.
  4. Kumar, Dinesh, et al. “A systematic study of double auction mechanisms in cloud computing.” Journal of Systems and Software 125 (2017): 234–255.
  5. Brenzikofer, Alain, and Noa Melchior. “Privacy-Preserving P2P Energy Market on the Blockchain.” 2019.
  6. Tushar, Wayes, et al. “Peer-to-Peer Trading in Electricity Networks: An Overview.” IEEE Transactions on Smart Grids, vol. 11, no. 4, 2020, pp. 3185–3200.
  7. BEUC — The European Consumer Organisation. Energy markets of the future: How the EU’s energy transition should work for consumers. 2017.
  8. Sousa, Tiago, et al. “Peer-to-peer and community-based markets: A comprehensive review.” Renewable and Sustainable Energy Reviews, vol. 104, 2019, pp. 367–378.
  9. Alliance for Internet of Things Innovation. “Open Energy Marketplaces Evolution — Beyond Enabling Technologies”. 2021.
  10. Z. Elmrabet, et al. “Cyber-Security in Smart Grid: Survey and Challenges”
  11. Balogh, Lilla Sarolta. “Az igazsággépezet: Michael J. Casey–Paul Vigna: The Truth Machine–The Blockchain and the Future of Everything St. Martin’s Press, New York, 2018, 302 o.” Közgazdasági Szemle 66.9 (2019): 1005–1009.
  12. Mühle, Alexander, et al. “A survey on essential components of a self-sovereign identity.” Computer Science Review 30 (2018): 80–86.
  13. Miyamoto, Toshiyuki, et al. “An energy trading system with consideration of CO 2 emissions.” IEEJ Transactions on Electronics, Information and Systems 125.10 (2005): 1513–1521.
  14. Gimpel, Henner, and Juho Mäkiö. “Towards multi‐attribute double auctions for financial markets.” Electronic Markets 16.2 (2006): 130–139.
  15. Nassiri-Mofakham, Faria, et al. “Bidding strategy for agents in multi-attribute combinatorial double auction.” Expert Systems with Applications 42.6 (2015): 3268–3295.
  16. Mengelkamp, Esther, et al. “Designing microgrid energy markets: A case study: The Brooklyn Microgrid.” Applied Energy 210 (2018): 870–880.
  17. Beiler, Kevin J., et al. “Architecture of the wood‐wide web: Rhizopogon spp. genets link multiple Douglas‐fir cohorts.” New Phytologist 185.2 (2010): 543–553.
  18. M. Mihaylov, S. Jurado, K. Van Moffaert, N. Avellana, and A. Nowe,´“NRG-X-Change — A novel mechanism for trading of renewable energy in smart grids,” in SMARTGREENS, 2014, pp. 101–106.
  19. Märkle-Huß, Joscha, Stefan Feuerriegel, and Dirk Neumann. “Contract durations in the electricity market: Causal impact of 15 min trading on the EPEX SPOT market.” Energy Economics 69 (2018): 367–378.
  20. Khuntia, Swasti R., Jose L. Rueda, and Mart AMM Van der Meijden. “Long-term electricity load forecasting considering volatility using multiplicative error model.” energies 11.12 (2018): 3308.
  21. Lindberg, K. B., et al. “Long-term electricity load forecasting: Current and future trends.” Utilities Policy 58 (2019): 102–119.
  22. Zhang, Chenghua, et al. “Peer-to-Peer energy trading in a Microgrid.” Applied Energy 220 (2018): 1–12.
  23. van der Veen, Reinier AC, and Rudi A. Hakvoort. “The electricity balancing market: Exploring the design challenge.” Utilities Policy 43 (2016): 186–194.
  24. Lüth, Alexandra, et al. “Local electricity market designs for peer-to-peer trading: The role of battery flexibility.” Applied energy 229 (2018): 1233–1243.
  25. Kruizenga, Richard John. Put and call options: a theoretical and market analysis. Diss. Massachusetts Institute of Technology, 1956.
  26. Jeriha, J., et al. “National balancing and wholesale electricity markets structure and principles. CROSSBOW Consortium.” (2019).
  27. Correa-Henao, Gabriel Jaime, and José María Yusta-Loyo. “Representation of electric power systems by complex networks with applications to risk vulnerability assessment.” DYNA, 2015.
  28. Holmgren, Ake J. “Using Graph Models to Analyze the Vulnerability of Electric Power Networks.” Risk Analysis, vol. 26, no. 4, 2006.
  29. Albert, Réka, Hawoong Jeong, and Albert-László Barabási. “Error and attack tolerance of complex networks.” nature 406.6794 (2000): 378–382.
  30. MIT Energy Initiative. “Utility of the Future — An MIT Energy Initiative response to an industry in transition,” 2016.
  31. Singh, Harmanjit, and Richa Sharma. “Role of adjacency matrix & adjacency list in graph theory.” International Journal of Computers & Technology 3.1 (2012): 179–183.
  32. Cormen, Thomas H., et al. Introduction to Algorithms. Third ed., The MIT Press, 2009.
  33. Jargstorf, Johannes, Kris Kessels, and Ronnie Belmans. “Capacity-based grid fees for residential customers.” 2013 10th International Conference on the European Energy Market (EEM). IEEE, 2013.
  34. Temple, William G., and Richard TB Ma. “Monotonic marginal pricing: Demand response with price certainty.” ACM SIGMETRICS Performance Evaluation Review 41.3 (2014): 65–70.
  35. Namerikawa, Toru, and Yoshihiro Okawa. “Distributed Dynamic Pricing in Electricity Market with Information Privacy.” Economically Enabled Energy Management: Interplay Between Control Engineering and Economics. Springer Singapore, 2020. 213–244.
  36. Borowski, Piotr F. “Zonal and Nodal Models of energy market in European Union.” Energies 13.16 (2020): 4182.
  37. Pérez-Arriaga, Ignacio J., and Yves Smeers. “Guidelines on tariff setting.” Transport pricing of electricity networks. Springer, Boston, MA, 2003. 175–203.
  38. Wang, Yishu, et al. “Time-dependent graphs: Definitions, applications, and algorithms.” Data Science and Engineering 4.4 (2019): 352–366.
  39. Meeus, Leonardo, and Nouicer, Athir. The EU clean energy package. 2018.
  40. G. C. Okwuibe, M. Wadhwa, T. Brenner, P. Tzscheutschler and T. Hamacher, “Analysis of Key Performance Indicators for Local Electricity Markets’ Design Analyse des indicateurs clés de performance pour la conception des marchés locaux de l’électricité”, in IEEE Canadian Journal of Electrical and Computer Engineering, 2021.

--

--

Engineering open source software that simulates and operates grid-aware energy exchanges, creating local marketplaces that interconnect to form a smart grid

Love podcasts or audiobooks? Learn on the go with our new app.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store