TSO-DSO Coordination in Decentralized Ancillary

Services Markets
Nikolaos Savvopoulos, Theodoros Konstantinou and Nikos Hatziargyriou
School of Electrical and Computer Engineering,
National Technical University of Athens (NTUA), Athens, Greece
Emails: [email protected], [email protected], [email protected]

Abstract—Distributed Energy Resources, located in the C. Parameters

distribution networks are becoming a key component of
the undergoing power system transformation. The proactive PiDA Day ahead net real power injection at bus i
management of these resources in the ancillary services markets, (generation - demand)
can influence positively the transition of the energy system, PiDA,T D Day ahead real power injected at bus i of the
leading to higher integration of renewable resources. This substation
paper investigates the operation of various ancillary services’
market schemes, aiming to increase the services provided by the ∆Pi Unbalance on net real power injection at bus i
resources located in the distribution network. The investigated (generation DA - generation RT - (demand DA -
market schemes are modeled using DC-optimal power flow, demand RT))
while focusing on services related to energy balancing and CiG Regulation bid located at bus i
congestion management. The cooperation of system operators
(transmission and distribution) is investigated to promote the Bl Susceptance of line l
active participation of distributed energy resources. The proposed PlF,max Active power flow limit of line l
formulation is demonstrated on a combined transmission and G
Pi − Downward regulation limit offered at bus i
distribution network, considering unbalances, driven by forecast G
error, in the renewable generation. Pi + Upward regulation limit offered at bus i
Index Terms—ancillary services, balancing, congestion
management, distributed energy resources, real time market, I. I NTRODUCTION
TSO/DSO coordination.
In recent years, global power sector is shifting towards more
sustainable and climate-friendly electricity production driven
N OMENCLATURE by decreasing technology costs and governmental policies. As
a consequence, the shares of electricity produced by renewable
A. Sets energy sources, mainly solar- and wind-based generation, have
S set of system operators increased substantially. At the beginning of 2019, the installed
N set of system buses (nodes) global capacity of solar photovoltaic (PV) plants reached 402
G set of controllable generation buses GW and of wind power plants 539 GW [1]. More specifically
in Germany, wind covered around 21% and solar PVs around
L set of system branches (lines)
9% of the electricity consumption in 2018 respectively [2], [3].
T set of substations (transformers)
In order to meet the CO2 reduction targets of the European
Ns subset of buses per system operator s ∈ S Union, it can be expected that wind and solar energy will
Gs subset of generation buses per system operator constitute a major part of the electricity mix in Europe.
s∈S The growing share of distributed energy resources (DER)
Ls subset of system branches per system operator in the distribution grid provides opportunities to use these
s∈S resources for the provision of services, not limited to the
Ts subset of substations per system operator s ∈ S distribution grid, but for the overall benefit of the entire power
NTD
s set of substation buses per system operator s ∈ S system. The procurement of ancillary services (AS) from DER
requires optimal coordination between the transmission system
B. Variables operator (TSO) and the distribution system operator (DSO) [4].
The transformation of distribution grids to active networks
PiG Upward or downward real power regulation with bidirectional power flow has increased the importance
offered at bus i of real time coordinated operation of the transmission and
δi Voltage angle at bus i distribution grids. The paper fits the timely and growing
PlF Real power flow of line or transformer l research initiative which explores, from different angles, the
PiT D Real power injected at bus i from the substation tighter interactions and cooperation between TSO and DSOs.
978-1-7281-1156-8/19/$31.00 c 2019 IEEE
Several centralized and decentralized coordination schemes of all grid areas and has access to all power plants. This
have been proposed lately in the literature [5]–[9]. case takes the optimal re-dispatch decisions and serves as
In [5], Caramanis et al. present a direct iterative based a reference for the comparison with the rest market models,
approach, incorporating transmission and distribution marginal namely a single AS market model (TSO leader), a local AS
locational prices towards the integration of distributed market model model (DSO leader) and a shared balancing
resources to the energy markets. A dual-horizon rolling responsibility market model.
scheduling model for active distribution system management A. Centralized AS market
is presented in [6]. Another hierarchical approach for the
coordinated economic dispatch of TSO and DSO using a This approach considers a single system operator who
convex AC optimal power flow model is presented in [7]. is responsible for controlling the whole system. This
The authors in [8], [9] explicitly focus on AS markets and coordination scheme resolves the imbalance by simultaneously
the coordination between system operators for balancing of accounting for transmission and distribution constraints.
the whole system. A detailed review of coordination schemes Transmission and distribution resources are dispatched
between local and central electricity markets is presented according to an integrated optimization of the entire system.
in [10], in which the commonly used operational processes This model corresponds to an ideal market model considering
between system operators with respect to market design and full monitoring and control of all the system resources by a
security of the system are extensively analyzed. single system operator and mainly serves as reference for the
This paper extends the previous work and investigates following cases, since it depicts the best possible scenario.
the advantages of the coordination between TSO and DSO The power system is modeled as a set of buses N, connected
in the daily system operation. The paper focuses on the by a set of branches L, with controllable generators located
procurement of ancillary services for balancing and congestion at a subset of G ⊆ N of the system buses.
The objective is to minimize the cost of reserve activation
management from resources located in the distribution
grid. The nature of these ancillary services (balancing and X S X
X
congestion management) may be provided by resources that min CiG PiG + CiG PiG (1)
are not necessarily directly frequency controlled and therefore i∈G0 s=1 i∈Gs
different coordination schemes and markets (centralized and subject to:
decentralized) are applicable. It is noted that utilization of (i) transmission constraints:
DER for voltage control and reactive power balancing in X
distribution networks is attracting increasing attention, but are Pi = (Bij (δi − δj )) ∀i ∈ N0 (2a)
not within the scope of this paper. j∈N0

The remainder of this paper is organized as follows: Section Pi + PiG

DA
+ ∆Pi = Pi ∀i ∈ N0 − N0T D
II presents the formulation of the investigated AS market (2b)
models. Section III demonstrates the illustrative results for 24 PiDA + PiG + ∆Pi = Pi + PiT D ∀i ∈ N0T D (2c)
hours on a combined transmission and distribution system and G G
Section IV presents the conclusions. Pi − ≤ PiG ≤ Pi + ∀i ∈ G0 (2d)
PlF = (Bij (δi − δj )) ∀l = (i, j) ∈ L0
II. M ODELLING OF TSO-DSO COORDINATION SCHEMES (2e)
The paper investigates the impact of TSO-DSO coordination − PlF,max ≤ PlF ≤ PlF,max ∀l ∈ L0 (2f)
in the AS market. We consider that the day-ahead (DA) market
clearing mechanisms have already calculated the cost minimal (ii) distribution constraints for every DSO s = 1 · · · S:
X
dispatch of power plants to meet the demand. Nevertheless, Pi = (Bij (δi − δj )) ∀i ∈ Ns (3a)
DA market is mainly an economic market and does not j∈Ns
necessarily consider all the constraints of the network such as
PiDA + PiG + ∆Pi = Pi ∀i ∈ Ns − NsT D
line and voltage constraints etc. The situation is exacerbated,
(3b)
considering the increased balancing requirement driven by the
potential forecast error of renewable generation and demand. PiDA + PiG + ∆Pi = Pi + PiT D ∀i ∈ NsT D (3c)
G G
The generation schedule from the DA market is handed over Pi − ≤ PiG ≤ Pi + ∀i ∈ Gs (3d)
the system operator who runs a real time market to calculate PlF = (Bij (δi − δj )) ∀l = (i, j) ∈ Ls
the minimal re-dispatch measures to prevent system constraints (3e)
violation using the updated forecast predictions. It becomes
clear that the cooperation of system operators is required to − PlF,max ≤ PlF ≤ PlF,max ∀l ∈ Ls (3f)
run the AS market. (iii) constraints at the point of interconnection between TSO
In this paper, we present four different market models and DSO:
promoting the coordination between system operators. The
first approach considers a centralized AS market operator, who PtF = PiT D = −PjT D ∀t = (i, j) ∈ T (4a)
is responsible for both balancing and congestion management −PtF,max ≤ PtF ≤ PtF,max ∀t ∈ T (4b)
 Network equations, both for transmission system (2a) and subject to:
distribution system (3a), are represented through linearized X
power flow equations. Pi = (Bij (δi − δj )) ∀i ∈ Ns (8a)
j∈Ns

B. Single AS market (TSO leader) Pi + PiG = Pi

DA
∀i ∈ Ns − NsT D (8b)
This scheme considers a single AS market, run by the TSO, PiDA + PiG = Pi + PiT D ∀i ∈ NsT D (8c)
who contracts directly ancillary services from DER connected G G
Pi − ≤ PiG ≤ Pi + ∀i ∈ Gs (8d)
in the DSO network. This coordination scheme varies from
PlF = (Bij (δi − δj )) ∀l = (i, j) ∈ Ls (8e)
the centralized AS Market since it dispatches the available
resources by ignoring the distribution network constraints. − PlF,max ≤ PlF ≤ PlF,max ∀l ∈ Ls (8f)
In order to avoid violation of the distribution network
The constraints referring to the power balance, as in (8b)
constraints, resources need to be prequalified, in the sense
and (8c), consider only the per-bus power injection (generation
that distribution resources are not offered in the AS market
or demand), forecasted at the stage of DA market. It is noted
if they may violate distribution network constraints. At the
that this problem needs to be solved before the clearing of the
stage of prequalification, if exists, the DSO validates the
TSO balancing market, and is therefore necessarily agnostic
participation of DER to the flexibility market, while respecting
about the actual realization of the real-power unbalance.
the local grid constraints. Nevertheless, in the current study we
The second step corresponds to an AS market cleared by
are not explicitly modelling the pre-qualification phase, since
the TSO for both congestion management of the transmission
extensive communication infrastructure is required between
system as well as balancing of the whole system (both
the system operators.
transmission and distribution). DER, located at the distribution
The single AS market is modelled as follows:
grid, can be offered to this market only after the clearing of
S X the local market by the DSO. DSO is responsible to aggregate
X X
min CiG PiG + CiG PiG (5) and transfer the remaining bids to the central market, assuring
i∈G0 s=1 i∈Gs only bids respecting its distribution network constraints are
considered.
subject to: The TSO market is modelled as follows:
(2a), (2b), (2c), (2d), (2e), (2f), (3d), (4a), (4b)
X X S X
X
PiT D = PiDA +PiG +∆Pi ∀s ∈ S−{0}, i = NsT D (6) min CiG PiG + CiG PiG (9)
i∈Ns i∈G0 s=1 i∈Gs

The constraints contain the transmission constraints for subject to:

generation and lines, the distribution constraints only for (2a), (2b), (2c), (2d), (2e), (2f), (3d), (4a), (4b), (6)
generation and the constraints for the substation flow limit. Gnew− Gnew+
Pi ≤ PiG ≤ Pi , ∀i ∈ Gs (10)
All the generation and demand unbalance in the distribution
network is reported by the DSO and is transferred to the The last equation introduces additional restrictions to
connection point as shown in (6). the flexibility resources located in the distribution network,
dictated by the DSO after the local market clearing.
C. Local AS Market (DSO leader)
This coordination scheme introduces a local market D. Shared Balancing Responsibility AS Market
operated by the DSO, promoting the active participation This scheme deviates from the local AS model by
of DSO in the AS market clearing. The local AS model transferring the balancing responsibility of each system to
activates resources depending on where the imbalance occurs, the corresponding system operator. Therefore, a predefined,
while shifting priorities towards the DSO. The DSO has the coming from the DA market, power exchange at the point
priority of contacting flexibility offered by DER located in of interconnection between TSO and DSO is established and
the distribution network. The TSO can only contract DER each DSO runs a local market for both local congestion
indirectly, after the DSO has aggregated these resources and management and energy balancing. In this coordination
has transferred them to the TSO AS market. scheme the TSO has no access to DER connected at the
The local AS market can be modelled as a two-step distribution grid. Therefore flexibility offered by the DER
optimization problem. The first step corresponds to a market is reserved exclusively for the DSO, in order to fulfill its
cleared by the DSO only for local congestion management. responsibilities with respect to local grid constraints and
More specifically, each DSO (s = 1 . . . S) solves the balancing.
optimization problem as follows. This scheme can be modeled as two subproblems
such that the TSO clears transmission-level imbalances
by using transmission-level resources only while DSO
X
min CiG PiG (7)
i∈Gs
clears distribution-level imbalances by using distribution-level
 2010 S 1010 100
D2 TN D1

Generation factor [%]

2009 2008 2007 2006 1006 1007 1008 1009
Bid 2:
80
2005 W -15 MW 1005 Forecast Observation
Bid 3: Wind TN
2018 S -15 MW 1018
4 3 60 Solar D1
2004 2011 1011 1004 Solar D2
2014
2017 Bid 6: 1017
-8 MW
1014 Bid 5: 40
+15 MW
2015 2013 2012 2003 2002 2001 5
2 1001 1002 1003 1012 1013 1015
20
Bid 7:
2016 +7 MW 1 Bid 4: 1016
+5 MW
Bid 1:
+20 MW 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Look ahead time [hrs]
Fig. 1. Combined transmission and distribution system
Fig. 2. Wind and solar forecast and observation

resources only. The optimization problem solved by each

TABLE I
system operator (TSO or DSO) s ∈ S, is modeled as follows: D ISTRIBUTION NETWORK LINE RATINGS
X
min CiG PiG (11) Line Rating [MW] Distribution lines
i∈Gs 100 1001 - 1002 1002 - 1003
1003 - 1004 1004 - 1005
subject to: 60 1005 - 1006 1006 - 1007
(3a), (3b), (3c), (3d), (3e), (3f) 1003 - 1012 1012 - 1013
1007 - 1008 1008 - 1009
PiT D = PiDA,T D i ∈ NsT D (12) 1009 - 1010 1004 - 1011
11 1013 - 1014 1013 - 1015
The objective function only involves resources located at 1015 - 1016 1015 - 1017
the same operational region and the power exchange between 1017 - 1018
system operators is fixed to a set-point, as expressed in (12).
Since the interface power exchange is fixed to the DA market
clearing result, the operations of the TSO are fully decoupled TABLE II
A NCILLARY SERVICES BIDS
from those of the DSO. The resulting balancing actions are
feasible, provided the local imbalances do not exceed the local Bid id Bus System Offer P [MW] Price [$/MW]
reserves. However, the resulting balancing action may be more 1 1 TN 20 60
costly than necessary, since the reserve resources cannot be 2 4 TN -15 -5
3 1005 D1 -15 -60
pooled.
4 1013 D1 5 80
III. I LLUSTRATIVE R ESULTS 5 1017 D1 15 100
6 2004 D2 -8 -40
A. Test system 7 2012 D2 7 15
We illustrate the benefits of the investigated coordination
schemes using MATPOWER [11] on a combined transmission
and distribution test system, as shown in Fig. 1. The network only consider balanced, single-phase system operation, but
is constructed using the 5-bus transmission system [12] and the framework can be extended to three-phase unbalanced
the 18-bus distribution system [13]. The distribution network networks.
is connected via HV/MV transformers to HV buses 2 and 5. The benefits of the coordination schemes are demonstrated
The line thermal ratings of the distribution system were not for one day divided in 24 hour intervals. We consider that all
contained in the original data and are adjusted as shown for D1 market participants (producers and suppliers) have submitted
in Table I. It is noted that similar thermal ratings are chosen their offers, based on their forecast, in the DA market which
for D2 as well. has been cleared to meet the demand with the minimal cost
In addition to the units available at the original network, we of generation. Following the real-time market, which is the
consider a 20 MW wind farm located at bus 4 of transmission main focus of this paper, minimizes the re-dispatch cost due
network, and two solar PVs, with a capacity of 5 MW each, to unbalances driven by forecast error and line congestion.
located at buses 1006 and 2014 of distribution network D1 and We assume forecast and observation for renewable
D2 respectively. A heavily loaded version of the distribution generation as shown in Fig. 2. The wind generation forecasts
system is chosen for D1, while an average loaded version is and observation refer to the wind farm connected to the
chosen for D2, corresponding to 100% and 70% of the demand transmission network. The scenario, in general, considers a
given in [13] respectively. general overestimation of the power production for the wind
Distribution networks differ from transmission grids in that farm, except of a few hours of underestimation during the
they are typically radially operated and have high R/X ratio. day and a high accuracy prediction at the beginning of the
Furthermore, the loading at the three phases is unbalanced day. Regarding the solar generation located at distribution
and the lines/cables are not transposed. In this work, we network D1, there is a trend of underestimating the power
production, exacerbated for 5 hours, possibly driven by a
 Balancing/ Congestion Req. [MW]
8 TN D1 Balancing Req. Constraint Violation
D2 Total 500 9
6 Single AS Local AS Shared Balancing AS

Unbalances [MW] 4 400 6

Cost of AS [$]
2
300 3
0
200 0
−2
−4 100 −3

−6 0 −6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Look ahead time [hrs] Hour

Fig. 3. Unbalances Fig. 4. Cost of Ancillary Services and Unbalances

wrong prediction for clouds during these hours. On the the congestion of the distribution lines. If there is no constraint
contrary, the forecast related to solar generation at distribution violation, as happening in the remaining hours of the day,
network D2 presents an overestimation in comparison with the the cost of ancillary services represents only the balancing
actual observations of generation. Although load demand is requirement and shows similar results to the single AS market,
variable and uncertain, it can be considered predictable due to since the balancing problem is solved by the market run by
the extensive investigation of load forecasting techniques over the TSO.
the past decades. Therefore, the unbalances occurred by load The shared balancing responsibility AS market model
forecasting are neglected. proposes a market scheme where each system operator is
responsible for both balancing and congestion management
B. Results using only resources located at its own network. The resulting
The total unbalances in MW that need to be covered in AS costs, as shown in Fig. 4, are higher compared to the
the AS market are depicted in Fig. 3 and are calculated as other two market models since the separate markets are fully
the difference between the forecast (used in the DA market decoupled and the reserve resources are not pooled.
clearing) and the observation (used in the real time market). The active power dispatch of each flexibility resource within
Positive values correspond to upward regulation (system the investigated market schemes is presented in Fig. 5. The
operator buys electricity), while negative values correspond to results in Fig. 5(a), which correspond to the centralized AS
downward regulation (system operator sells electricity). Apart market, serve as a reference, since they represent the optimal
from the energy balance, the AS market may offer congestion solution. For several time points at the results of Fig. 5(a), we
management services, since DA does not necessarily consider note an simultaneous activation of both upward and downward
the whole transmission and distribution system constraints. regulation. At these time points, the system operator requires
The cost for AS for each hour of the day is presented upward regulation to tackle the distribution line congestion and
in Fig. 4. In the single AS market, the TSO can contract downward regulation for the balancing of the whole system.
directly ancillary services from all available sources, regardless During the hours that there is no line congestion we
the system they are connected (transmission or distribution). identify similar solution layouts between centralized, single
Therefore, the market result can achieve the most economic and local AS market, since a single system operator has the
point, reaching a total cost of 833.7 $, as shown in Table III. responsibility of balancing the whole system (transmission and
Nevertheless, in this coordination scheme the market operator, distribution). On the contrary, in local AS market scheme,
namely the TSO, is not considering the constraint violation of shown in Fig. 5(c), and during the hours that lines are
the distribution grid. For the specific day a total amount of congested, the generator located in bus 1017 increases its
10.87 MW needs to be redispatched or curtailed during the generation set-point to limit the overloading of distribution
real time operation of the network. The cost of this energy to line 1013-1015. Depending on the sign of balancing and
be covered is difficult to be estimated and compared to the AS congestion requirements, we identify an increase of the
market cost, since it might affect critical residential loads, but balancing requirement in case they are of opposite signs like
it can surely be assumed that it will be way higher than the in hours 9-11, or a decrease in case they are of the same sign
AS market cost. like in hours 12-19.
In the local AS, the DSOs are clearing a market for The active power dispatch of the AS resources for the
local congestion management and then the TSO is clearing shared balancing responsibility market model are shown in
a market for transmission system congestion management as Fig. 5(d), indicating higher utilization of all the AS resources.
well as balancing of the whole system. DER, located in the This is mainly driven by the balancing requirement that
distribution network, are participating in both markets; directly enforces each system operator to balance its system using only
in the local AS market of their distribution network and local resources. Additionally, in several moments during the
indirectly in the TSO AS market as an aggregated bid by day we identify both upward and downward regulation at a
the DSO. In Fig. 4, local AS market indicates higher cost, subsystem, since line congestion enforces activation of upward
especially during 8-19 hour. This increased cost is driven by regulation and balancing of the subsystem enforces activation
 1017 (D1) 1017 (D1) 1017 (D1) 1017 (D1)
8 1005 (D1) 8 1005 (D1) 8 1005 (D1) 8 1005 (D1)
1 (TN) 1 (TN) 1 (TN) 1 (TN)
6 4 (TN) 6 4 (TN) 6 4 (TN) 6 4 (TN)
2012 (D2) 2012 (D2) 2012 (D2) 2012 (D2)
AS Dispatch [MW]

AS Dispatch [MW]

AS Dispatch [MW]

AS Dispatch [MW]
4 2004 (D2) 4 2004 (D2) 4 2004 (D2) 4 2004 (D2)

2 2 2 2
0 0 0 0
−2 −2 −2 −2
−4 −4 −4 −4
−6 −6 −6 −6
−8 −8 −8 −8
0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24
Hour [hrs] Hour [hrs] Hour [hrs] Hour [hrs]

(a) Centralized AS (b) Single AS (c) Local AS (d) Shared Balancing

Fig. 5. Active power dispatch of flexibility sources for the investigated AS market models

TABLE III and the transmission and distribution line thermal rating.
C OST OF A NCILLARY S ERVICES Single AS market does not consider the distribution network
constraints and therefore cannot guarantee a feasible solution.
System Centr. AS Single AS Local AS Shared Bal AS
TN 1801.6 833.7 714.1 3455.5 Local AS market achieves the most efficient results which are
D1 - - 1087.5 2634.0 comparable to the one provided by the centralizes AS market
D2 - - 0 72.0 model. Finally, assuming adequate reserves in the distribution
Total 1801.6 833.7 1801.6 6161.5 system, shared balancing responsibility model will not violate
physical constraints, however it appears to be the least efficient
solution.
of downward regulation. These are considered as the main
reasons for the increased additional cost for ancillary services R EFERENCES
within this market scheme, as shown in Table III [1] REN 21 Steering Committee, Renewables 2018, Global Status Report.
The general performance of the three investigated AS Renewable Energy Policy Network for the 21st Century, 2018.
[2] Fraunhofer ISE. (2018) Energy charts. [Online]. Available:
market schemes shows that the single AS market can https://www.energy-charts.de/
achieve the most economic point but cannot guarantee [3] H. Wirth, Recent Facts about Photovoltaics in Germany. Fraunhofer
the prevention of distribution line constraint violation. Institute for Solar Energy Systems ISE, 2018.
[4] M. Birk, J. Caves-Avila, T. Gomez, and R. Tabors, “TSO/ DSO
For the cases that distribution lines are not congested, coordination in a context of distributed energy resource penetration,”
local AS market model may achieve results of the same in MIT Energy Initiative Reports, 2017.
economic performance, but extensive communication between [5] M. Caramanis, E. Ntakou, W. W. Hogan, A. Chakrabortty, and
J. Schoene, “Co-optimization of power and reserves in dynamic
the various system operators is required. Finally, the T&D power markets with nondispatchable renewable generation and
shared balancing responsibility model has limited real-time distributed energy resources,” Proceedings of the IEEE, vol. 104, no. 4,
communication requirements and this is depicted in the pp. 807–836, April 2016.
[6] A. Saint-Pierre and P. Mancarella, “Active distribution system
high operational AS cost for both balancing and congestion management: A dual-horizon scheduling framework for DSO/TSO
management. interface under uncertainty,” IEEE Transactions on Smart Grid, vol. 8,
no. 5, pp. 2186–2197, Sep. 2017.
IV. C ONCLUSION [7] Z. Yuan and M. Reza Hesamzadeh, “Hierarchical coordination of
TSO-DSO economic dispatch considering large-scale integration of
Derived by the necessity of developing local electricity distributed energy resources,” Applied Energy, vol. 195, pp. 600 – 615,
markets to increase the procurement of AS, we investigate 2017.
in this paper four types of AS market models, focusing on [8] N. Natale, F. Pilo, G. Pisano, G. G. Soma, M. Coppo, R. Turri,
G. Petretto, M. Cant, and G. Gigliucci, “Assessment of price and quantity
coordination schemes between DSO and TSO. The main of ancillary services provided by active distribution systems at the
goal is to increase the procurement of AS by facilitating TSO/DSO interface,” in CIRED Workshop 2016, June 2016.
distributed energy resources (DERs), flexible consumers and [9] A. Papavasiliou and I. Mezghani, “Coordination schemes for the
integration of transmission and distribution system operations,” in 2018
prosumers in distribution grids. The ideal market model of Power Systems Computation Conference (PSCC), June 2018.
a centralized AS market controlling the whole system is [10] Y. Tohidi, M. Farrokhseresht, and M. Gibescu, “A review on
compared to a single AS market, where the TSO activates coordination schemes between local and central electricity markets,” in
2018 15th International Conference on the European Energy Market
the DER for balancing and transmission system congestion (EEM), June 2018.
management, to a local AS market scheme, which promotes [11] R. D. Zimmerman, C. E. Murillo-Sanchez, and R. J. Thomas,
the participation of the DSO, and to a shared balancing “Matpower: Steady-state operations, planning, and analysis tools for
power systems research and education,” IEEE Transactions on Power
responsibility market scheme, where each system operator is Systems, vol. 26, no. 1, pp. 12–19, Feb 2011.
responsible for balancing and congestion management of its [12] F. Li and R. Bo, “Small test systems for power system economic studies,”
own system. The problem is formulated as a DC optimal in IEEE PES General Meeting, July 2010.
[13] W. M. Grady, M. J. Samotyj, and A. H. Noyola, “Minimizing network
power flow adopting cost-based objective function to identify harmonic voltage distortion with an active power line conditioner,” IEEE
the most economically beneficial plan, redispatching the DA Transactions on Power Delivery, vol. 6, no. 4, pp. 1690–1697, Oct 1991.
clearing results, while respecting the balancing of the system