Continuing our series on where and how, in our opinion, the electric power industry of Kazakhstan should evolve, this time we will focus on the market infrastructure - the electric grid complex, which should be organized in a coherent manner in order to build an efficient energy market that is resilient to external shocks.
Let’s say right away that Kazakhstan has a lot to do in this area. In fact, the country’s power grids are almost not organized in terms of any market approaches, and this immediately became noticeable as soon as in July 2023, they tried to introduce something similar to an power market in Kazakhstan in the form of a centralized Single Purchaser model. Suddenly it turned out that the bulk networks, managed by the System Operator KEGOC, created many years ago and, at first glance, seemingly an established attribute of the market infrastructure, in fact, are literally hanging in the air in terms of the market architecture and corresponding functionality, and are in no way connected with distribution networks in the consumers’ tariff structure where KEGOC services are charged in separate payments.
In the model of prevailing bilateral agreements that dominated the country until recently, this was acceptable; the “routing” of electricity transmission services was created in accordance with the locations of production and consumption. Upon the introduction of mandatory centralized trading for generation unaffiliated with large consumers, the issue of correctly constructed network infrastructure architecture and correct tariff setting, which most accurately reflects the cost structure of networks, based on topology and interaction with each other, as well as reserves in the power system as a whole, has become more relevant than ever. Disputes have already begun between energy suppliers and electric grid companies regarding payment for the latter’s services when trading through a Single Purchaser, related precisely to the uncertainties of the power system architecture and tariff setting principles. Taking into account the presence of significant volumes of self-generation among large consumers not integrated into the general market model with the SP, but at the same time having been using the country’s power system, these problems will only increase if its architecture and approaches to setting tariffs are not brought into line with the canons of the market.
With the transition to the SP model, tariffs for KEGOC services of the backbone networks use were introduced for subjects of the wholesale market, and separate transmission tariffs for importers and consumers with their own generation. They continue to be paid separately from distribution network tariffs. Charging for the use of KEGOC networks through a volumetric rate in kWh for all participants in the wholesale market (which KEGOC itself appoints) means that every kWh produced and consumed in the energy system through the SP system is paid in favor of KEGOC, regardless of the degree of participation of the main networks in the transmission given volume of electricity. Indeed, consumption around the same specific CHP thermal plant in a given location on the wholesale market is related to the use of KEGOC networks only indirectly in terms of reservation, while consumption from this thermal power plant in another region directly loads these networks, but consumers of the wholesale market in both cases pay the same. On the other hand, consumers outside the mandatory pool and importers pay a completely different tariff, significantly higher than the user fee, and again in kWh, although KEGOG costs depend on the transmission volume only in terms of losses. In general, differentiation of consumers based on energy supply through a common mandatory pool or from their own sources destroys the unity of regulation in the power system and its architecture and, of course, in the future, during the transition to a competitive market, must be eliminated.
In a competitive market model, the structure of the energy system must comply with market principles of correct costs reflection and distribution between market participants in accordance with their contribution to certain costs borne by certain network organizations.
In general, the structure of relationships in a modern power system is as follows. Bulk networks transmit electricity from large power plants to distribution networks, to whose substations consumers are connected. Part of the generation, for example cogeneration (CHPP), some power plants of large consumers, are not directly connected to the backbone networks (substations 220 kV and above), and are essentially distributed sources, since their electricity is mainly consumed within the radius of the effective heat supply of the CHPP or at a short distance large consumers, transmitted to them through distribution networks (and sometimes directly from power plant buses). There is also other distributed generation close to consumers, either in the distribution networks of the power system, or “behind the meter” directly in the consumer networks. Such generation participates in the energy supply process either directly or in an aggregated form (for example, virtual power plants).
Consumers of such generation (or the generation itself) in any case use the main networks, since the distribution networks are connected to the power system, and this means that the system provides backup to these consumers (and even to the distributed sources themselves - thermal power plants, etc.) which in some modes use the services of the backbone networks directly, while in normal modes they use it as a reserve. This means they must pay for these services. To some extent, the wholesale consumers tariff for using KEGOC, introduced in connection with the transition to the SP model, is an attempt to solve the problem of paying for reservations in the power system, but it applies to the wholesale market entities appointed by KEGOC itself and is transmitted onto all of them, regardless to the extent of the backbone grid use (since it is assumed that all consumers receive electricity from the Single Purchaser), and this is its main imperfection. Such a payment creates a cross-subsidy by consumers of CHP plants and distributed sources using only system backup, in favor of consumers who actually receive electricity from wholesale power plants. In addition, the tariff for using the power system in a certain part must be paid by consumers with their own generation, which, just as in the case of distributed heat and power plants, are reserved by the power system.
An equally pressing problem of the current Kazakh regulation is that the tariff of bulk networks is formed exclusively in volumetric (volume) indicators, in kWh, while the services of these networks consist of both the direct transmission of electricity and the maintenance of networks in readiness for transmission and reservation, therefor KEGOC’s costs for operation, and especially for investments, depend little on transmission volumes.
In the mandatory pool mode with a virtual electric power supply from the SP, this is all the more obvious; the service fee in this case does not directly depend on the volume of electricity pumped through the main grid. And this problem needs to be solved today, and even more so, it must be kept in mind when modeling a competitive market in the country. As a rule, in the world, for the sake of simplicity, tariffs for backbone networks are expressed through a single rate for content - a postal stamp - and a rate for losses - that is, a two-rate tariff.
The architecture of distribution network tariffs is formed somewhat differently. These are the tariffs that mass consumers, both wholesale and retail, encounter. In Kazakhstan, we observe uniform electricity grid tariffs of distribution companies for all groups of consumers, expressed in volumetric indicators - tenge per kWh. This approach grossly ignores the basic and universally applied principle of differentiation of tariffs for electricity transmission - according to voltage levels. It is quite obvious that the lower the voltage level at which consumers are connected and pay for transmission services with power grids, the higher the costs of power grids and, accordingly, the lower the costs of consumers who need to spend less in both operating and capital costs. A plant that consumes electricity at the voltage of 110 kV must have its own high-voltage power grids, appropriate operational services, etc., that is, incur significant costs. And for a store or persons dwelling in an apartment or at a household that receives electricity at a voltage of 380/220 V, these types of costs tend to zero. Therefore, the electricity grid tariffs of Kazakhstan, with such regulation, already include significant volumes of cross-subsidization of the population, budget, small and medium-sized enterprises on the part of larger consumers with their own electrical equipment - volumes of cross subsidies that for some reason no one remembers when discussing the problems of cross subsidies of tariffs for guaranteeing suppliers contrasting the independent ones. In most countries, especially more or less large ones, regardless of whether their electricity sector is regulated or market-based, tariffs are somehow differentiated by voltage levels, and this was even the case in the USSR, when the population paid 4 kopecks/kWh, and industry 2 kopecks.
With market approaches to organizing tariffs, differentiation of electric grid tariffs by voltage levels is mandatory since it reflects the main principle of market pricing - cost allocation. But not only. Let's consider the cost structure of electric grid companies and their possible reflection in the tariff, which is the most important principle of correct tariff formation - cost reflection.
The first component, reflecting the actual transmission of electricity through transmission losses up to the blocking of the heating factor which, when using a nodal market model, is expressed through a change in the marginal price at a given network point, the so-called congestion price (that is, the rate for compensation of losses in the network, reflecting the actual transfer of electrons from point A (generation) to point B (consumption) of the network, in the limiting case equal to the difference in prices between the point of electricity generation in the network and the point of consumption).
In distribution networks, the nodal model is not yet used, so the rate for losses is calculated on average, without congestion price, based on the network parameters (the higher the voltage, the lower the current, and, accordingly, losses, plus wire materials, etc.) and is expressed in costs per kWh.
The second component from top to bottom is the cost of providing peak power, or maximum network capacity. The electrical network can be associated with a road, which must be designed to accommodate as many vehicles as possible during rush hour. This component can be reflected in the tariff as a rate for peak power in money per kW. But the network capacity should always be higher than the maximum possible peak power, because some operational and balance reserve is always needed, otherwise we will have to shed some consumers at some point, without ensuring short-term reliability, which in many countries is directly normalized through probabilistic indicators. That is, the network should always be a little “expandable,” just like the road, which during construction should take into account not current, but forecast traffic. But both the network and the roads cannot be built and expanded in small parts, in increments, there is always an abrupt expansion and strengthening - new substations, power lines, equipment, redundancy, etc., including at a higher level - in backbone networks, therefore There are always costs that cannot be compensated either by the loss rate or by paying for peak power, no matter how accurately they are calculated and correlated with their demand by consumers. And these are the so-called residual costs, which are very difficult to distribute among consumers other than according to their criterion of using the network - both power and transmission itself. And these are the most socialized costs, which in a market scenario are best reflected through the rate for having a connection to the network, expressed through the power of the consumer’s electrical receivers - that is, his potential for consuming both power and electricity.
These three components of network costs can be displayed through two components in kWh and kW for peak power (which, in fact, is usually done in most countries for large consumers), and even through one - only in kWh, as is the case in Kazakhstan, where on top of that it also involves averaging over voltage levels.
But even with two rates, and especially one, we are faced with a violation of the principles of cost allocation (correct attribution of costs) and causation (causality - the one who creates the costs pays). At one rate, those who consume little, but always have a reserve from the network (these are large consumers with their own generation or distributed resources – DERs behind the meter), essentially shift the costs of the networks onto those who consume a lot. With two rates, it is possible to implicitly compensate for the third rate - in the rate for capacity, include residual costs: reserve, advance of network development “for growth”, etc., and thus transfer them to consumers during network power control hours, but at In this case, on the demand side there is always the opportunity to consume less during these hours, and thus engage in free riding in relation to network tariffs, without paying sufficiently for the reservation from the power system, shifting the costs to other consumers. In addition, the explicit display of residual costs in the tariff through the rate for the presence of connection, firstly, gives a clear idea of the quality of regulation (a too high rate signals excessive investment costs in the network complex and reserves, and too low indicates their insufficiency), and secondly, it allows you to directly link the costs of technological connection with tariffs for transmission services, reducing the often prohibitively high payment for networks connection, transferring its payment into time-extended payments for the connection availability.
Ignoring these factors not only creates injustice and cross-subsidization in the market, the need to increase tariffs for everyone, but also distorts economic signals in the network complex in terms of correctly reflecting the cost structure through network tariffs, hinders the development of distributed energy resources, creates a conflict between own generation or, more broadly, consumer resources and the power system. Such conflicts have been existing in many countries, especially in connection with the development of low-carbon DERs, the introduction of various benefits to consumers in terms of technological connection to the network, when consumers end up overordering the power they need in the power system many times, and the substations built for them, which are unprofitable for the networks, stand unloaded with unclaimed transformers, and losses are ultimately transferred to those who consume a lot and constantly, and whose demand is inelastic, that is, small and medium-sized businesses, budgets and the population.
When developing a market scenario for the development of the electric power industry in Kazakhstan, it is necessary to take into account all this negative experience and draw conclusions - to design tariffs in distribution networks in such a way as to maximally reflect the costs of networks in the architecture and structure of tariffs. These networks require a three-rate tariff, at least for large and medium-sized consumers, at which they will pay for losses, peak power, and the costs of networks and the power system as a whole for reservation and advanced development to current demand through the rate for the connection availability.
The rate for the connection availability can also be used to optimize the payment for reservation by the power system of consumers with their own generation, who use distribution networks in one form or another, which is relevant for Kazakhstan. Indirectly, this rate should partially take into account the rate for the maintenance of the KEGOC tariff, which will be transmitted not through the tariff for the use of the bulk grid to all subjects of the wholesale market in the form of some kind of quasi-tax payment, regardless of their grid usage degree and expressed in kWh, as is the case today, but through distribution networks, through their rate for the availability connection of consumers (and possibly generators), which they in turn must partially pay to KEGOC in accordance with the capacity of consumer facilities connected to these networks. The tariff rate for the connection availability may also include various payments related to regulatory policies, for example, cross-subsidization of vulnerable segments of the population, or business contributions for additional support for renewable energy sources, that is, what is called policies. Considering that we are talking about regional networks, and even in some cases intra-regional - municipal, policies can be highly localized, and therefore precise in application. The question is to ensure that regulators do not abuse this tool, so the distribution of residual costs should correspond to the so-called. Ramsey's principle, the essence of which is to impose costs in inverse proportion to the elasticity of consumer response - the richer consumers are and the less responsive they are to additional costs, the greater the share of such costs they bear. In the conditions of Kazakhstan, it is obvious that these should be large holdings and enterprises; for them this will be a relatively small, but significant burden for society and the region.
Thus, the tariff architecture will take on a harmonious structure - all distribution networks directly (if they border on main networks) or indirectly (if they border through other distribution networks) in their tariff have a rate (or rates, considering the rate for losses) of the KEGOC tariff, which they pay according to actual power to KEGOC. At the same time, the source of these payments are payments for peak power and for the connection availability to the network from consumers. The loss rate is based on the volume of flows from KEGOC networks and own losses in power distribution networks.
We will talk about how transmission tariffs in Kazakhstan should be formed in a market scenario, and how this relates to competition in the market, in our next article in this series.