It is now widely recognised that hydrogen, blue or green,¹ may make a substantial contribution to achieving "net zero" greenhouse gas emissions. Hydrogen can be used as a fuel in heating and transport, and in industrial processes, or as a vector for transportation and/or storage of electricity.

In the UK, government and industry are considering how to make this happen, and a number of pilot projects are in development. It is thought that early large-scale hydrogen production projects would be developed to supply existing demand of industrial customers, located close to the hydrogen production, seeking to decarbonise their processes. Another (perhaps complementary) route to market for hydrogen would be injection into the GB natural gas transportation system, where it would be blended with natural gas and supplied to consumers as part of their gas supply.

In September 2020, Cadent and Frontier Economics published a report (the blending report) on Hydrogen Blending and the Commercial Framework, which addresses blending of hydrogen and the issues to which it would give rise under the commercial framework for the gas transportation system.

A separate report (for BEIS) published by Frontier Economics in August 2020 on Business Models for Low Carbon Hydrogen (the funding report) discusses possible forms of financial support for hydrogen production (funding support), primarily in the context of industrial customers, but also referring to hydrogen blending as a possible "back-up" or "spill" route to market where demand from industrial customers falls short for any period.² It is recognised that the commercial framework for blending and the regime for funding support are closely related.

A core focus of the blending report is the safe limit on the proportion of hydrogen which can safely be blended with natural gas in the system (without the need, for example, to modify consumer appliances). A 20/80 blend (hydrogen/methane, by volume) is currently considered the upper limit. The blending report is clear that hydrogen blending within such limit will be only a temporary and transitional arrangement, towards a hydrogen-only (or hydrogen plus biomethane) system (if that becomes an ultimate policy goal).

The blending report recommends that, at least initially, it would not be worthwhile pursuing complex but "perfect" solutions for the transitional commercial framework (for example, because blending limits may rarely apply in practice), and that a pragmatic approach should be taken in some areas – making minimum necessary changes without regrets. However, this approach will need to allow for the fact that investors in hydrogen production projects may want relatively long-term stable rights in terms of system access, as well as funding support. Within the life of a supported project, more complex circumstances may in future develop which would require more elaborate solutions. The question arises whether these solutions could be pre-defined with enough clarity (and certainty for investors) to allow them to be "switched-on" when needed but not earlier.

This paper looks at certain recommendations of the blending report and starts to consider how those recommendations might be implemented in the context of the Uniform Network Code (UNC) which governs the commercial framework referred to in the report. Understandably, promoters of hydrogen projects will be waiting for clarity on the government's intentions on funding support, but there may be merit in considering (or even initiating) possible modifications of the UNC in advance of that, and perhaps pushing forward the discussion of some of the associated wider policy issues. This paper assumes some knowledge of the UNC on the part of the reader; it includes (at the end) a glossary of UNC (and other) terms.

We are grateful for Frontier Economics for their review of this paper, which resulted in several improvements.

In this paper, we look at three areas addressed in the blending report: the trading of hydrogen blended in the system (and gas balancing); the allocation and pricing of capacity in the system for hydrogen injection; and the management of hydrogen blending limits. Some of the possible approaches discussed would fall into the "future, complex circumstances" category mentioned above.

Other issues recognised but not addressed in the blending report (nor in this paper) include the impact of hydrogen blending on levels of shrinkage in the system, and on flow weighted average calorific value (FWACV) billing to consumers. 

A theme of the blending report is the increased costs which hydrogen injection blending may cause for the system. For any such cost, there may be a choice between:

• specifically identifying and recovering it (from the person injecting hydrogen), and then allowing for it in the level of hydrogen funding support available; and
• leaving it to be borne by all users of the system, and ultimately consumers. This would be the default position, absent changes in the commercial framework of the UNC, assuming that the price controls of the transporters (the network operators) allow those costs to be passed on.

If funding support were consumer rather than taxpayer backed,³ there may be little to choose between these in terms of practical outcome.

Commercial arrangements for sale of hydrogen

As a preliminary matter, the blending report assumes that the blending process begins with the injection of unblended hydrogen, rather than following a requirement of first blending hydrogen with methane to a required specification prior to injection.⁴

The blending report envisages that arrangements for gas trading and gas balancing would not need to change to accommodate the injection and blending of hydrogen.

From a regulatory perspective, this means that someone would need to "arrange with a gas transporter for [the hydrogen] to be introduced into" that transporter's pipeline system,⁵ and so be licensed as a gas shipper (unless exempt), and a Shipper User under the UNC. The shipper (hydrogen shipper) could be:

• the hydrogen producer itself (selling the hydrogen at the NBP);
• an existing gas shipper buying the hydrogen (as part of its own gas balance) from the producer (or possibly an on-site customer of the producer) at or before the point of entry to the system.

In either case the implication is that, with funding support, hydrogen can be sold at a price (p/kWh) which is competitive with natural gas, taking account of system entry costs which (as discussed below) may differ from natural gas.⁶

A contract for difference (cfd) against the wholesale (NBP) natural gas price would be a possible support mechanism to achieve this, at least for blended hydrogen, and perhaps for industrial customers for which it displaces natural gas as a fuel. In such a cfd, the strike price (stated at the NBP) would need to cover the cost of hydrogen entry capacity (discussed below) as well as hydrogen production. If the combination of industrial demand with blending provided a sufficiently secure source of demand for hydrogen, a simple, single-priced, commodity cfd could be workable, without having to split pricing with a capacity-type payment to cover fixed costs.

Hydrogen entry capacity

In UNC parlance, the hydrogen shipper would deliver hydrogen to the Total System at a System Entry Point. The blending report differentiates between injection at the distribution level (to an LDZ) and the transmission level (to the NTS).

Injection at distribution level

At the LDZ level, the blending report recommends a requirement⁷ that capacity in the LDZ should be held by the hydrogen shipper, and priced on a long run marginal cost (LRMC) basis, recognising that increasing levels of LDZ injection connections (for both biomethane and hydrogen) may mean that existing levels of LDZ capacity are insufficient for future flows. This arrangement would replace the existing LDZ system entry commodity charge applied to biomethane projects.

We note the following points:

• Unlike the NTS (for which entry and exit capacity are separately defined and held), in the LDZ only a single class of capacity (LDZ Capacity) is currently defined. Shipper holdings of LDZ Capacity are determined in respect of exit flows from the LDZ. The DN Operator is responsible for entry flows into the LDZ at NTS/LDZ Offtakes and it would currently be pointless to require the DN Operator to hold its own entry capacity.
• The hydrogen entry capacity would be a new class of entry capacity under the UNC (and not LDZ Capacity as currently defined, although physically both are capacity in the LDZ). New definitions and rules for allocating the entry capacity would be required in TPD Section B. Its pricing would presumably take account of the pressure tier within the LDZ at which the entry point is located (for exit capacity, the size of maximum exit flow is used as a proxy for this).
• Presumably for level-playing field reasons, all kinds of injection to the LDZ would be required to hold this LDZ entry capacity, including biomethane and potentially any LDZ-connected storage sites.
• If the entry capacity is to be priced on a basis which recognises LRMCs, it should presumably take account of the difference in physical pipeline capacity required for transportation of hydrogen compared to that required for natural gas (reflecting the differences in calorific value and other characteristics). This implies (on the assumption that the capacity is defined in energy units, i.e. kWh/day, like all other capacity under the UNC) that there would be different classes of capacity for hydrogen and natural gas (or a multiplier applied to the capacity charge at hydrogen sites).
• As noted above, this higher cost of entry capacity for hydrogen would need to be accounted for in any funding support mechanism.
• Query whether applying the LRMC methodology would require an (implicit or explicit) allocation of LDZ entry capacity to the DN Operator at NTS/LDZ Offtakes (although from a UNC drafting point of view that could be avoided by defining the new entry capacity as applying at points of entry to the Total System on the LDZ).
• Once in the system (entry-paid on the basis of this new LDZ entry capacity) we assume there is no continuing differentiation of hydrogen – it is simply accounted for as energy for balancing and other purposes. The customer, and the Shipper User at LDZ exit points (and the amount of LDZ Capacity the shipper holds), and indeed the trader at the NBP, is indifferent to the composition of the blend.⁸ However, the difference in physical transportation characteristics of hydrogen and natural gas means that physical LDZ Capacity all the way down to the LDZ exit points is used more intensively, which is presumably taken into account in the LRMC determination of the price of hydrogen entry capacity.
• Query whether LDZ entry capacity would (initially or eventually) be tradable (LDZ Capacity is not, as it is allocated effectively on an administered basis).

Injection at transmission level

For hydrogen entry to the NTS, the blending report proposes no change in current arrangements. The hydrogen shipper would apply for or acquire and hold NTS Entry Capacity, undifferentiated in definition or price. This is based on Ofgem’s conclusion, in approving a postage stamp methodology,⁹ that (at least for the time being) the absence of constraints in NTS capacity means that there is no need to differentiate costs by location and no distortions to competition in not doing so.

Blending constraints

As noted above, a key focus of the blending report is the hydrogen blending limit.

As a preliminary point, setting a blending limit assumes the indifference of consumers to the blend composition up to the limit. The report does raise the issue of end-users that might require a specific gas composition (for example, a minimum 85% methane) for a given industrial process or power generation technology. This could be a significant complication, but probably only likely to arise on the NTS. It might be addressed by (re-)configuring pipework – for example, to inject hydrogen downstream of the end-user’s connection, or abstract natural gas upstream of the hydrogen connection, or by installation or adaptation of end-user plant. The question would then arise as to who should bear the cost and, linked to that, whether existing end-users have an (indefinite) expectation as to gas composition. We have not addressed this issue further in this paper.

The report identifies that, when a hydrogen plant connects to the system at a given point and injects, it (as it were) consumes blending capacity within the system. It limits the extent to which another plant can inject hydrogen downstream of the entry point and, conversely, it could be constrained in its ability to inject by another hydrogen plant upstream of it.

The report recognises two aspects of blending:

• the long-term issue of allowing connections and "allocating" blending capacity to hydrogen projects seeking connection;
• the short-term issue of managing injections of hydrogen within the blending limits.

These two aspects interact – if hydrogen entry capacity were fully interruptible, then it could, in theory, be allocated without regard to the impact on and of other hydrogen producers.

Long-term issue – allocation of blending capacity

On the long-term issue (allowing connections/allocating blending capacity):¹⁰

• Although conceptually different, blending capacity and entry capacity would presumably be allocated in a single process – the allocated entry capacity would be the absolute limit on hydrogen injection (unless both could be held on an interruptible basis).
• The report proposes an "administered" approach – an assessment would be made of the impact of connecting a new production facility at a given location on other (existing or potential) hydrogen productions and, if the assessed impact was adverse to some defined degree, then connection would not be allowed at that location.¹¹
• It is worth considering whether some kind of market-based allocation mechanism could work without being over-complex. Periodically, the "blending capacity" at different locations on the system could be assessed and published. There would be published window periods within which applications for connections would be considered (each effectively a blending capacity allocation round).
• Applications would be evaluated against some commercial (as well as technical) criteria. Once allocated, the capacity would have a defined degree of firmness (in terms of entitlement to possible compensation – see below).
• Whatever blending allocation mechanism is used, there is a sequencing question with the allocation of funding support. Applicants for funding support would be expected to have developed their projects to a certain degree before they can apply (to screen out purely speculative projects, and allow the project to be costed) and that could include having secured connection rights. Possibly, where projects proposing to connect along the same path (and therefore competing for blending capacity) are also competing for funding support, the allocation of blending capacity could be suspended until the outcome of the funding application was known; or a connection offer could be required to be surrendered or transferred if the project does not succeed in its funding application.
• If (for a particular producer) blending is only used as a back-up where (in a given timeframe) industrial consumer demand falls or fails, that would add a further uncertainty in forecasting the extent of blending constraints (perhaps depending how similar are the customer characteristics for different producers).
• It may be necessary to define "available blending capacity", at least for some allocation mechanisms. This would raise questions about "technical capacity". In other words, at what probability or frequency (of the blending limit being met) would blending capacity at a given point be defined? Would that be based on the "status quo" (current configuration of the network and level of connected load), or would it allow for planned future developments. Furthermore, would the allocation of blending capacity come with a promise not to erode it in future by allowing new upstream connections? These issues are tied to the issue of compensation for being constrained off – see below.

Short-term issue – dispatch

On the short-term issues (constraining hydrogen injection to the blending limit):

• As noted above, the blending report envisages this may be a rare occurrence, at least initially, and overcomplicated solutions may not be worthwhile; but (on the other hand) investors will want certainty over the period of (at least) their contract for funding support.
• An initial point in the report is whether this would be self-administered (the hydrogen producer would constrain itself, based on flow composition analysis) or centrally (with the relevant system operator "dispatching" off the producer). This is a second-order issue – the system operator would need a fall-back ability to shut out injection if the producer did not do so itself.
• Where more than one producer contributes to the blend in any part of the system, there may be a choice as to which is constrained off. That decision could, for example, be "last-in first-out", or downstream before upstream, or pro rata capacity or injection nomination, or something else.

There are difficult questions about the financial impact on the shipper/producer of being constrained off:

• Producers could be expected to price in the constraint risk to the required level of funding support. That might only be appropriate (from the perspective of the taxpayer/consumer) if the extent of constraints could be forecast with some accuracy. But where blending is a back-up to industrial consumer demand, the constraint risk could be seen as second order.
• The blending report notes that a funding support mechanism might need to address the constraint risk. That could entail, for example, a cfd with a split tariff.
• A different approach would be for the constrained producer to be compensated under the UNC framework (in other words, by the system). Assuming such compensation came within the scope (under the UNC) of transporter "neutrality" to costs of balancing and constraint management costs, the effect would likely be similar (to consumer-funded funding support), in that the costs would ultimately be passed to the consumer.
• Compensation could be at a pre-determined level (set when the capacity and/or funding support was awarded, possibly as a bid parameter). Alternatively, it could be set by some kind of market-based process on the day, if the blending constraint could be solved by action taken by more than one producer, with different cost consequences. For example, some producers may access alternative demand on the day, or have access to hydrogen storage. Liquidity in any such arrangement would only develop when there is a significant number of hydrogen connections along a given path.
• Compensation by the system (for hydrogen production being dispatched down or off) could take the form of an energy buy-back (which would keep the hydrogen shipper's gas balance unimpaired) or a capacity buy-back (which would not).
• Any of these compensation mechanisms risks being expensive for the taxpayer/consumer, and possibly over-compensating the producer (although the mechanism should take account of avoided costs (for example, of feed-gas for blue hydrogen).
• A possible compromise could be to define a future level of occurrence of constraints at which a compensation mechanism would be switched on (which might be never). The producer would take the constraint risk up to that point. Principles of compensation could be set so as to assure the producer of a certain level unless a market-based approach (to be developed in future) solved the constraint more cheaply.

Other issues to be considered in the context of dispatch to meet blending constraints include:

• The necessary information flows – which may be complex (including flow rates and blend compositions at different points and times along a given path, and possibly alternative network configurations).
• The extent to which forward notice of a constraint can be given (allowing a hydrogen producer to take action to mitigate the cost consequences/access alternative hydrogen demand).
• Whether a system operator could manage a constraint (or adhere to a forward constraint notice) by accessing within-day natural gas storage and linepack variation (and the costs of doing so).
• Whether the information flows and "dispatch instructions" are centralised through National Grid NTS or (at the LDZ level) managed by individual DN Operators.
• At the drafting level, whether the rules are set out in the UNC, or in each project's Network Entry Agreement and, linked to this, whether the shipper or the hydrogen plant operator is responsible for implementation.

Indicative UNC modifications

To allow for hydrogen blending, aside from other parts of the regulatory framework, modifications of the UNC (mainly the Transportation Principal Document) would be needed, including in the areas discussed above.

An initial consideration is whether to make these in each relevant UNC section, or to group them together in a separate new component document of the UNC (e.g. a "Hydrogen Blending Document"), operating to vary specific provisions of the TPD where required for hydrogen blending. The advantage would be to keep the variations in one place, rather than scattering them across the code. That was the approach taken in relation to implementation of the European network codes (by the creation of the European Interconnections Document).

Whichever approach is taken, some of the required changes in TPD would be (in outline):

Section A – System Classification

• Define a hydrogen entry point as a subset of System Entry Point (for both NTS and LDZ). This would probably be necessary to allow for distinguishing hydrogen entry capacity and the application of different entry rules (see B and I below).

Section B – System Use and Capacity

• Define LDZ entry capacity as a new class of capacity. Specify the basis of application for and allocation of this capacity (for all categories of entry – including from storage, biomethane as well as hydrogen). Define a new LDZ entry capacity charge and specify the basis of charging (and remove the existing LDZ entry commodity charge).
• For hydrogen, as discussed above, the allocation rules would need to address blending impacts and possibly the interaction with funding support allocation – these rules could sit in a separate document (referenced in Section B) whose ownership and governance might flow from hydrogen-specific regulation and involve the Secretary of State/Ofgem as well as the transporter.
• If any future commercial framework for "dispatch" to manage blending constraints was based on "capacity buy-back", the rules would be specified in Section B.

Section D – Operational Balancing and Trading Arrangements

• If such a future commercial framework for dispatch was based on "energy buy-back", the rules might be specified in Section D (possibly, such that a constraint management action would be treated as a Trade Nomination, in which case in principle no change might be needed in Section E (Daily Quantities, Imbalances and Reconciliation).

Section F – System Clearing, Balancing Charges and Neutrality

• The cost of energy-based actions to manage blending constraints would be added in the list of items counted in energy balancing neutrality.

Section I – Entry Requirements

• Significant new rules would be required over the combination of Section I and a hydrogen Network Entry Agreement. The exact allocation of rules between those documents would be driven in part by the need to bind the hydrogen producer directly to operational rules (such as for constraint management), and partly by drafting convenience.
• Changes in Section I may include:
  • Define a hydrogen production plant as a new subset of Connected Delivery Facility, and possibly define a hydrogen Delivery Facility Operator separately.
  • Define new scope of Network Entry Provisions and possibly amend the scope of Gas Entry Conditions.
  • Consider whether a hydrogen SEP can be combined with other SEPs in a single ASEP.
  • Address interaction of blending constraint management with capacity constraint compensation rules.
• New rules in a Network Entry Agreement may include:
  • Separate entry specification for hydrogen.
  • Modifications to measurement provisions if needed for hydrogen.
  • Rules and procedures for blending constraint management.
  • Local Operating Procedures (LOPs) modified to reflect the above.

Section J – Exit Requirements

• The Applicable Offtake Arrangements will need to allow for a blend including hydrogen to be made available for offtake at System Exit Points. That will presumably follow (through the definition of Standard Offtake Arrangements) from an amendment or a relaxation of the Gas Safety Management Regulations.
• Any arrangement to meet a specific composition requirement for a particular end-user could then involve a Special Offtake Arrangement.

Section Y – Charging Methodologies

• Set out the new LRMC methodology for LDZ entry capacity.

Other UNC sections

• Other changes could be required in Sections N (Shrinkage), O (System Planning), Q (Emergencies) and S (Invoicing and Payment).

1. Hydrogen produced by electrolysis using a renewable power source is described as "green"; hydrogen produced from methane with carbon capture and storage, as "blue".↩
2. From the perspective of funding support, blending is not seen as a "high value" use of hydrogen compared with displacing carbon-intensive use of industrial end-users.↩
3. For example, if it were similar to the renewable cfd model administered by LCCC.↩
4. The report notes that in Germany in some cases a "pre-mix" is required.↩
5. Section 5(1)(c), Gas Act 1986↩
6. The position would be different if the basis of funding support were a purchase obligation, by gas shippers or possibly the network operator, but that approach is not preferred in the funding report.↩
7. This is a change from the current position where shippers delivering gas to the LDZ are not required to hold entry capacity.↩
8. The report notes the significant work to be done in respect of the FWACV arrangements in this context.↩
9. UNC modification 0678A, for implementation of the EU gas tariff code.↩
10. We have assumed that only a single shipper would deliver hydrogen and hold entry capacity for a given production project. That might not be true (for example, on a project expansion), but multiple shippers could add complexity.↩
11. Or allowed only on terms as to interruption which would make it commercially unviable.↩

Glossary of terms

cfd: contract for difference

DN Operator: Distribution Network Operator

FWACV: flow weighted average calorific value – basis on which calorific value of gas supplied can be determined for customer billing

LDZ: Local Distribution Zone, a geographic division of a distribution network

LRMC: long run marginal cost

National Grid NTS: National Grid Gas plc as operator of the NTS

NBP: National Balancing Point, the notional point at which gas is traded in the wholesale GB gas market

NTS: National Transmission System

NTS/LDZ Offtake: a point at which an LDZ is connected to the NTS

SEP: System Entry Point

Shipper User: a licensed gas shipper as User under the UNC

Total System: the total system comprising NTS and LDZs

Transporter: National Grid NTS or one of the DN Operators

TPD: Transportation Principal Document

UNC: Uniform Network Code