Energy grids

Question 1

why are energy grids relevant?

Answer 1

for the transport and distribution of energy

Question 2

how is electricity transported?

how is gas transported?

Answer 2

Electricity

• transport via cables and overhead lines.
• transmission and distribution need different levels
• typically national grids (transmission) and local grids (distribution), national grids linke by interconnectors

Gas

• transport via pipelines or ships
• distribution on different pressure levels
• pipelines are often international, grids local

Question 3

why is storage always at the customer’s end of the pipeline?

Answer 3

• > To ensure no shortage, smooth demand.

- >smaller and cheaper pipelines due the storage in each region

Question 4

what are some responsibilities of the TSO?

Answer 4

=>reliable network
=>maintenance
=>enlargement
=>buy renewable electricity
=>system restoration after blackout
=>frequency control
=>voltage control

Question 5

how many DSO has Germany?

what are their responsibilities?

Answer 5

about 1000

same responsibilities as TSO (except from frequency control and system restoration)

Question 6

what is the UTCE

Answer 6

Union for the coordinaton of transmission of electricity,

estab. in 1951 by 8 countries (BEL, GER, IT, FR, LX, SZ, AU). Each partner is responsible for the permanent electricity supply in his control area.

Question 7

Which current dominates when transmitting and distribution?

Answer 7

AC

Question 8

In which case HVDC (High voltage direct current) is used

Answer 8

long distances, undersea cables, interconnection with non-synchronised network systems (e.g. Sweden)

Question 9

what is natural monopoly and why are grids natural monopolies?

Answer 9

Natural monopoly: production of a good/service by a single firms ensures lowest cost (ie. monopoly is the most efficient way to organise the market - from the perspective of the production cost).

Typically when the marginal cost are constant or decreasing in business with long investment / long innovation cycles (not a lot of innovation)

Grids distributing energy show decreasing marginal cost within the area of the grid (strong economies of scale - adding more houses to the grid is cheap), and less when enlarging or merging. => IOW: Its cost inefficient to build and operate 2 electricity or gas grids in parallel in the same area, consequently energy grids are natural monopolies.

Question 10

What is a direct network effect?

Answer 10

Direct network effect:
the value of a good/service increases as the number of users of the product/services increase ( telephone )

Positive network externality
Value increase for the individual user after a new user access.
->this creates positive feedback loop

Question 11

what is a two-sided network effect?

what is the positive loop for two-sided network effects

Answer 11

value of a good/service increase for the individual as number of of users of complementary products increase ( indirect network effect)

new A users join => value to B users increase => New B users join => value to A users increase => new A users join….

Question 12

electricity grids show typically two-sided network effects. explain who are the two sides an how positive two sided networks effects can appear?

Answer 12

grid operators and customers
-> increase on customer makes the company more reliable => positive two sided networks effects further increase the economic benefits of grid monopolies, strengthening the monopoly.

Question 13

what are the 2 problems with grid monopolies? what are their respective solutions?

Answer 13

Monopolistic pricing = output is restricted and prices are higher than in an efficient market. Pricing is particular inefficient when elasticity of demand is low.
=>Solution: 1. price regulation to create market like results 2. state ownership (however high incentive for raise prices (vs. collecting taxes))

Vertical integration of grids with retail or generation/production (one single company operating the grid, selling electricity and operating the power plant)
This can result in significant market power in retail or whole sale markets (discrimination against competitors using the grid) ; consequence: less competition in wholesale and retail, inefficient market
=>Solution: regulated third-party access, separation of grid business from other business (unbundling)

Question 14

what is the change of paradigm?

Answer 14

third party access
legal unbundling of transmission and distribution network
>organisational, legal and accounting unbundling
>Chinese walls (informational unbundling)
>independent regulator in each member state

Question 15

how is the new structure of the energy markets in Germany?

Answer 15

transmission and distribution businesses are separated from other businesses and their prices are regulated by Bundesnetzagentur.

Question 16

what are the consumers, products and pricing of the grid?

Answer 16

Consumers or retail org. offering the bundle of energy provision and transport via grid

Products: reliable transport of electricity via the grid.

Pricing: electricity grid is an indivisible good > direct allocation of cost to single transports is challenging: for all flows all available power lines are used!

Thus uniform prices for energy delivered and or peak load, to be paid by the consumer depending on his consumption or peak load (because thicker cables for peak load are more expensive)

Question 17

what are the goals of grid regulators ?

Answer 17

To find a balance:
> interest of consumer are low prices (no mon. pricing)
and quality grids (few black outs)
>grid companies need reliable, predictable income to find investors for their businesses and the state wants to incentivise sufficient new investment in grids.

Question 18

what is the traditional approach for the rate of return?

Answer 18

Revenues = OPEX + Depreciation + RAB * r

• OPEX: operating expenditures are fully refunded (if you spend high you get high but no extra profit)
• Depreciation: Depreciated capital invested by the operator at replacement cost (comparable to capital employed)
• RAB: regulatory asset base (the more you invest the higher the RAB > more extra profit)
• r: allowed rate of return on the capital invested (set by regulator)

Question 19

what the problems with the rate of return regulation?

Answer 19

if r is lower than for equally risky investments in the capital markets > no new investments in the grid.

As the ideal r is impossible to determine, it will be set at level attractive for investors > grid operator has incentive to maximise his profits by setting the ratio of Capex to opex inefficiently high, increasing the RAB (over investing =gold plating). This incentive to over-spend in Capex is called Averch-Johnson-effect

for the remaining open, the grid operator has no incentive to minimise cost, i.e. to be really efficient, as all reasonable costs are compensated (problem of information assymetry between the operator and the regulator)
=> Rate of return regulation can lead to inefficient grid operators.

Question 20

What are alternative approaches to the rate of return?

Answer 20

The incentive regulation:
create incentives for the grid operator to lower cost (move to the status that would be the result of competition in a functioning market)

Question 21

what are the 2 approaches of the incentive regulation?

Answer 21

• > modest incentive approach: offering the operator a benefit for disclosing his inefficiencies: the regulated revenue is set like in a rate of return regulation but valid only several years (regulatory period). If during that period actual cost fall below the regulated cost, the extra profit goes to the operator. At the beginning of the new period the extra profits are removed by a new regulated revenue cap.
• > strong incentive approach (yardstick): set a grid operator’s revenues based on an avg. cost of the industry in total. Every operator who is more efficient than his peers can make extra profit as long as he is more efficient. Others make losses.

Question 22

how is the modest incentive implemented?

and what are the problems?

Answer 22

=> Price cap vs. Revenue cap: cap (set by the regulator) on total revenue or on individual prices for services

=> RPI - X regulation: cap on price or revenue does not remain constant during the regulatory period. It is annually adjusted by an annualised growth factor g (+/-)

gt = RPIt - Xt

RPI = retail price index (proxy for cost increases due to infla)
X = expected productivity improvement

Problems:

• not fair for historically efficient firms as X is universal (firms that already efficient when system started are the one who are punished - not fair)
• weak incentive tu cut cost at the end of the regulatory period
• incentive to reduce the quality of the grid

Question 23

how can be the strong incentive regulation implemented?

Answer 23

RPI - X regulation combined with benchmarking:

• calculate the relative efficiency of the individual company based on other’s companies data ( X is thus determined by the company’s efficient; an inefficient company will have a high X and will have pressure to low cost) (100%efficient > X=0 you and gt increase > can increase your cap)
• the benchmark can be avg performance or best practice
• used in Eu countries.
• after benchmark short- term adjustment of all operator’s revenue according to each operator’s measured efficiency compared to avg efficiency.

Question 24

why might be benchmarking (strong incentive regulation) inefficient?

Answer 24

technical inefficiency
scale inefficiency
low capacity utilisation after investments
non-accounted quality differences in inputs/outputs
environment: climate local market
uncertainty or measurement error

Question 25

why do we use benchmarking (strong incentive regulation)?

Answer 25

in order to estimate cost efficiency, defined as the ration of actual cost to minimal cost to proceed an equal or higher amount of products and services under equal or more server circumstance ( actual cost > minimal cost => inefficient)

Question 26

how is the benchmarking procedure?

Answer 26

• Input (total cost), Output (bundle of product and service parameter -cable length and area serviced)
• Determine meaningful indicators (ratios) that express aspect of the cost efficiency of a grid operator (e.g. cost per km of cable in the grid);
• Calculate all efficiency ratios for all grid operators based on one year

Question 27

parameters necessary to compare efficiency (Electricity)

Answer 27

Electricity:
Customer service
-number of metering points per voltage and transformer level- houses connected to the grid
-number of connection points per voltage level
-lenght of cables and overhead lines

Capacity provision

• peak load per voltage and transf. level MW
• delivered area
• installed decentralised generation capacity

Transportation Work
-total delivered electricity per voltage level - important bus has loses when transporting (1%)
-

Question 28

parameters necessary to compare efficiency (GAS)

Answer 28

Customer service

• number of metering points per pressure level
• number of number connecting points per p. level
• length of pipelines

Capacity provision

• peak feed and outage rates
• delivery area
• population serviced

Transportion work
-total delivered gas volume

Question 29

what are the resulting ratios of the parameters

Answer 29

area (m^2) / annual cost of the grid (EUR);
metering points/ cost;
peakload / cost;
cable length/cost

Question 30

what are the benchmarking approaches?

Answer 30

DEA (Data envelopment analysis) benchmarks each company against real peers that are structurally similar.

Linear programming calculation.

Non parametric: weak a priori assumptions.

Deterministic: all data is considered to contain significant information (weakness)

SFA (Stochastic Frontier Analysis) benchmarks each company using a function the is derived from data of all other companies.

Regressive calculation

Parametric: cost function defined a priori expect for some unknown parameters that are estimated from data

stochastic: tries to separate random noise from significant information

Question 31

what are the strengths and the weakness of the DEA and SFA?

Answer 31

DEA (suitable for companies hard to compare and has strengths with small set companies but needs reliable data)
strengths:
-makes a weak a priori assumptions
-handels multiple inputs
-benchmarks agains real peers (company is only inefficient if similar, more efficient companies exist)
-cautions evaluations (typically results in high efficiency scores > easy to defend in court)

weakness:

• cannot handle statistical noise in data (risk of data error)
• Outliers can massively influence results

SFA (best suitable for large samples of rather homogeneous companies with potentially noisy data, where statistical meths can play out their strengths)
strengths:
-Strong theory of significance testing
-separates noise and efficiency
-handles multiple inputs (Dimensions)

weakness

• requires large data sets
• requires strong a priori assumptions (risk of specification error)

Question 32

what are the challenges with today regulatory approaches? what are some possible solutions?

Answer 32

Quality: conventional grid regulation focus on cost > incentives to neglect quality (under-invest). Possible solution: quality premiums (however hard to measure quality - long time lags)

Self supply: because grid tariffs depend on energy consumption, their cost however depend on peak load consumption. This creates incentives for consumers to supply themselves while staying connected to the grid (they pay thus less). Thus the cost remain unchanged and the revenues will decrease, and as a consequence the tariff will go up for other consumers. This will again create incentives for customer to supply themselves making the tariffs even more higher. (Vicious circle)
Solution: Tariffs based on capacity (reduce the capacity that you take of the grid and this will reduce the cost of the grid)

Incentives for generation:
As generation does not pay grid tariffs, there is no incentive to build up new generation where the grid “needs it”. New decentralised generation can need massive investment in distribution grids to be borne by the local consumers.
Possible solutions: tariffs for generation depending on site; however complex to implement

capacity constraints in a world of intermittent, decentralised electricity production: These constraints can appear for short periods in different part of the grid
Possible solution: nodal pricing, i.e. grid tariffs differentiated by time and location, depending on current capacity utilisation in each part of the grid.