Phenomena of Parallel Operation aspects of the MTDC Systems
❖ The major motivation to design Multi-Terminal Direct Current
(MTDC) systems is the integration of large-scale offshore power
sources, such as wind turbines and solar systems, to meet the need for
flexible, reliable, long-distance electricity transfer.
❖ In a parallel MTDC system, multiple converter stations are connected in
parallel, maintaining a constant DC voltage across all terminals while
allowing individual current control for power sharing and lower
losses. This configuration enables staged development, facilitates power
reversal, and is suitable for integrating multiple sources and loads.
❖ Parallel MTDC systems allow for n-1 redundancy, meaning even if one
terminal fails, others can continue operation. This enhances system
reliability and availability.
❖ The parallel operation of converters within MTDC systems introduces
several physical phenomena and technical challenges, necessitating
sophisticated control strategies and protection mechanisms to ensure
system stability, power sharing, and fault flexibility.
Key Phenomena and aspects associated with their Parallel Operation
❖ These systems, employing multiple voltage source converters (VSCs) interconnected
through a common DC grid, are particularly well-suited for large-scale renewable energy
integration, intercontinental power exchange, and secure, efficient long-distance
transmission. Parallel operation means operating multiple converters or terminals
simultaneously connected to the same DC grid. This introduces several operational
phenomena like;
Mitigation/Control
Aspect Phenomenon Method
Droop control, load dispatch
Power Sharing Unequal load/generation sharing
algorithms
Master-slave voltage control,
Voltage Stability Instability due to uncoordinated control
decentralized droop
Fast DC breakers, protection
Fault Propagation Grid-wide collapse on fault
zones
Synchronization, impedance
Circulating Currents Voltage mismatch between terminals
design
Control Complexity in managing multiple Centralized or decentralized
Coordination terminals control
Technology limitations (e.g., current Proper converter rating and
Converter Capability
limit) protection
Key Phenomena and aspects associated with their Parallel Operation @Continuation
1) DC Voltage Control and Stability
▪ In MTDC systems, only one terminal (or a few) can control the DC
voltage, while others operate in power control mode. Parallel operation
demands a voltage-control hierarchy, typically: One voltage-controlled
converter (master) and Others in current/power control (slaves). Any
voltage deviation must be handled quickly to avoid instability.
▪ Only one or a few terminals operate in voltage control mode (regulating
𝑉𝑑𝑐 ), while others operate in power control mode. The controlled terminal
sets the DC grid voltage, whereas others inject/absorb power accordingly.
Voltage deviation affects system stability, governed by:
𝒅𝑽𝒅𝒄
𝑪𝒅𝒄 = 𝑰𝒊𝒏𝒋 − 𝑰𝒍𝒊𝒏𝒆
𝒅𝒕
Where:
▪ 𝐶𝑑𝑐 : Equivalent DC capacitance;
▪ 𝐼𝑖𝑛𝑗 : Current injected by converter;
▪ 𝐼𝑙𝑖𝑛𝑒 : Current flowing in the DC lines.
This differential equation shows the dynamic behavior of the DC voltage under changing
conditions.
Key Phenomena and aspects associated with their Parallel Operation @Continuation
2) Load Flow and Dynamic Interaction
▪ Power flow is governed by the DC line impedances, terminal
voltages, and control setpoints. Parallel operation can lead to
circulating currents if there is a mismatch in voltage setpoints
between converters. Accurate modeling of line resistance and
converter dynamics is essential.
▪ Inaccurate synchronization of voltage levels between converters
can cause circulating currents. This is especially a problem in
parallel-connected VSCs. The circulating current between two
terminals is approximately:
𝑽𝒅𝒄𝟏 − 𝑽𝒅𝒄𝟐
𝑰𝒄𝒊𝒓𝒄 =
𝑹𝒍𝒊𝒏𝒆
Where 𝑅𝑙𝑖𝑛𝑒 is the DC line resistance between terminals. Hence, even
small voltage mismatches can lead to significant circulating losses.
Key Phenomena and aspects associated with their Parallel Operation @Continuation
3) Fault Handling and Protection
▪ A fault at any point in the DC grid affects all connected terminals.
In parallel operation, fast fault isolation is critical to prevent
cascading failures. The use of DC circuit breakers and fast
coordination algorithms is necessary.
▪ In MTDC systems, faults (like DC short circuits) propagate
rapidly due to low line inductance and the absence of natural zero
current crossings. The system requires fast fault detection and
isolation using DC circuit breakers. Energy during a fault is:
1 2
𝐸𝑓𝑎𝑢𝑙𝑡 = 𝐶𝑑𝑐 𝑉𝑑𝑐
2
This stored energy must be dissipated or diverted, necessitating
robust protection schemes.
Key Phenomena and aspects associated with their Parallel Operation @Continuation
4) Control Coordination Strategies: For stable operation, converters
must coordinate through either centralized control or Decentralized
control. In the case of centralized control, one controller governs
the whole grid (less robust), whereas in decentralized control, each
converter makes local decisions based on droop characteristics (more
scalable).
An emerging approach is hierarchical control, which combines
primary (droop), secondary (voltage restoration), and tertiary
(economic dispatch) layers to balance performance and resilience.
5) System Modeling: The stability of parallel operation can be assessed
using eigenvalue analysis of the linearized system, ensuring all
eigenvalues lie in the left half of the complex plane for stable
behavior. The dynamic behavior of MTDC systems under parallel
operation is often studied using small-signal modeling. The system
stability is evaluated through eigenvalue analysis of the linearized
state-space representation.
Control of Power Flow in MTDC Systems
❖ The control of power flow in a Multi-Terminal DC (MTDC) system ensures
reliable operation, proper power sharing among terminals, and stability of
the entire DC grid. Power flow in MTDC is managed primarily through
voltage source converters (VSCs) and their control strategies.
❖ In AC grids, flexible alternating current transmission systems (FACTS) are
used to control active and reactive power. However, in DC grids bus
voltages are only characterized by their magnitude, and the impedance of the
bus doesn’t contain any imaginary part. Hence, for an MTDC control, only
magnitudes are used to control active and reactive powers.
❖ In MTDC systems, power flow between terminals depends on the voltage
difference and line impedance:
𝑉𝑑𝑐,𝑖 (𝑉𝑑𝑐,𝑖 − 𝑉𝑑𝑐,𝑗 )
𝑃𝑖𝑗 =
𝑅𝑖𝑗
Where: 𝑃𝑖𝑗 is power from terminal i to terminal j,
𝑅𝑖𝑗 is the resistance of the line connecting terminals i and j
𝑉𝑑𝑐,𝑖 and 𝑉𝑑𝑐,𝑗 are the DC voltage at terminal i and j.
Control of Power Flow in MTDC Systems @ Continuation
❖ To manage this power flow, MTDC systems utilize Voltage Source Converters
(VSCs), which can operate in different control modes: voltage control, power
control, or droop control. Voltage control mode is typically used at one
terminal (called the slack or reference terminal) to maintain the DC voltage of
the system, while the other terminals operate in power control mode to inject or
absorb specific amounts of power. A common and effective method for
decentralized power sharing in MTDC systems is droop control, which links the
terminal’s output power to its DC voltage using the relation
𝒓𝒆𝒇 𝒓𝒆𝒇
𝑷𝒊 = 𝑷𝒊 − 𝑲𝒊 (𝑽𝒅𝒄,𝒊 − 𝑽𝒅𝒄 )
𝑟𝑒𝑓
Here, 𝐏𝐢 is the active power injected by the 𝒊𝒕𝒉 terminal; 𝑃𝑖 is the reference
𝑟𝑒𝑓
power; 𝑉𝑑𝑐,𝑖 is the local DC voltage; 𝑉𝑑𝑐 is the nominal DC voltage; 𝐾𝑖 is the
droop coefficient.
❖ The droop control method allows converters to autonomously adjust their power
contribution based on local voltage measurements, without the need for fast
communication links.
❖ Thus, through a combination of voltage regulation, power references, and
droop-based adjustments, MTDC systems achieve flexible and reliable control
over power flow in large-scale interconnected networks.
Control of Power Flow in MTDC Systems @ Continuation #Repeat
❖ Power Sharing and Control Mechanism
▪ Ensuring that all terminals share power proportionally based on their
capabilities. Use of droop control or centralized dispatching algorithms
to coordinate power injection/withdrawal without overloading a particular
terminal.
▪ In parallel operation, it is essential that the terminals share the DC power
appropriately. The most common method to achieve this is through droop
control, which relates the power injection 𝑃 with the DC voltage 𝑉𝑑𝑐 :
𝒓𝒆𝒇 𝒓𝒆𝒇
𝑷𝒊 = 𝑷𝒊 − 𝑲𝒊 (𝑽𝒅𝒄,𝒊 − 𝑽𝒅𝒄 )
Where:
▪ 𝐏𝐢 is the active power injected by the 𝒊𝒕𝒉 terminal;
𝑟𝑒𝑓
▪ 𝑃𝑖 is its reference power;
▪ 𝑉𝑑𝑐,𝑖 is the local DC voltage;
𝑟𝑒𝑓
▪ 𝑉𝑑𝑐 is the nominal DC voltage;
▪ 𝐾𝑖 is the droop coefficient.
This linear relation ensures that terminals adjust their power based on the voltage
deviation, allowing multiple terminals to share power without explicit
communication.