Undervoltage load shedding (UVLS) disconnects blocks of load in stages when the voltage in an area of the grid stays low for several seconds, a sign that the reactive power demand exceeds what the network can deliver. Its purpose is to prevent voltage collapse. Unlike underfrequency load shedding, it acts on the affected area and must tell a sustained undervoltage from a sag caused by a fault.
Why sustained undervoltage calls for load shedding
The voltage in an area of the grid is held up by reactive power. When the reactive demand exceeds what the sources and lines can deliver, for example after losing a line or a generator in a heavily loaded area, the voltage starts to fall. If nothing stops it, the process feeds itself: induction motors draw more reactive power at reduced voltage, the on-load tap changers of distribution transformers raise the voltage on the load side and restore its consumption, and the transmission network sags further. The result can be a voltage collapse within seconds or minutes.
Undervoltage load shedding (UVLS) is the last automatic resort: it disconnects blocks of load in the affected area so that demand falls back within what the network can sustain.
How it differs from underfrequency load shedding
| Underfrequency load shedding | Undervoltage load shedding | |
|---|---|---|
| What it indicates | Active power deficit in the whole system | Reactive power deficit in one area |
| Reach | Global: frequency is the same across the system | Local: voltage varies from one bus to another |
| Speed | Stages in fractions of a second | Stages in seconds, to ride through fault sags |
| Where load is shed | Spread across the system | In the area losing voltage |
The frequency-based scheme is explained in underfrequency load shedding. In a plant with its own generation, load shedding follows a different logic, described in load shedding in plants.
How a UVLS scheme is built
A typical scheme has several stages. Each one combines a voltage threshold, a time delay and a block of load:
- The first stages act at moderately low voltages with long delays, so that load is not shed for a dip the grid could correct by itself.
- The following stages act at lower voltages with shorter delays, because the collapse is closer.
- The delays must always be longer than the fault clearing time, including that of backup protection.
The measuring element is an undervoltage 27 element at the substation bus, which trips the breakers of the feeders assigned to each stage. Thresholds, delays and blocks come from a voltage stability study of the real system; there are no universal values.
There are two architectures:
- Decentralized: each substation measures its own voltage and sheds its own feeders. It is simple and robust.
- Centralized: a controller receives the voltage of several buses and decides which loads to disconnect across the area. It allows more selective criteria, at the cost of depending on communications.
Blocking and supervision
- Is the voltage below the stage threshold on all three phases?
- YesHas it stayed there for the stage time delay?
- YesIs there a voltage circuit failure or an ongoing fault detected by overcurrent?
- YesBlocked: no load shed
- NoTrips the feeders of the stage
- NoNo trip: it was a sag
- NoNo trip
- Three phases: requiring undervoltage on all three phases avoids tripping for unbalanced faults, which depress only one or two.
- Voltage circuit failure: a blown VT fuse looks like an undervoltage. The scheme must be blocked, as explained in loss of potential alarm.
- Ongoing fault: some schemes are blocked while overcurrent elements detect a fault, so that no load is shed during the sag.
- Delayed voltage recovery after a fault: with a large share of motor load, voltage can take several seconds to recover after the fault is cleared. That is precisely one of the cases UVLS must handle, so the overcurrent block must release once the fault is cleared.
Restoration
Shed loads do not come back by themselves. Restoration is manual or authorized by the system operator, because reconnecting before reactive support is restored can repeat the collapse. If the substation has automatic restoration logic or automatic reclosing, the shed feeders must be excluded from it.
Regulatory framework
In North America, NERC PRC-010 sets a framework for the design, assessment and operation of UVLS programs, with obligations for planners and for the entities that own the equipment. In other countries, the system operator decides whether an undervoltage scheme is required, in which areas and on what criteria. Where it is not required, a utility or a plant may have its own local scheme to protect its network.
What goes wrong in practice
- Shedding during sags, because delays are shorter than backup fault clearing.
- Tripping on a VT failure with no blocking.
- Stages that never act because they require a threshold the voltage never reaches before the collapse.
- Wrong load blocks: feeders reassigned without updating the scheme, or essential loads included by mistake.
- Automatic restoration of shed feeders.
How it is tested
In general terms, each stage is checked with secondary injection: threshold on all three phases, time delay, tripping of the assigned feeders, blocking on voltage circuit failure and on overcurrent, and exclusion from automatic restoration. Periodic tests and the records of each operation are part of the scheme’s maintenance.
We can help with your load shedding scheme
At Tiza Group we carry out the load flow and stability study that defines the stages through our engineering design and studies service, and the remote relay configuration of the shedding logic and its blocking, for substations and industry.
Related guides
- Undervoltage Protection 27 (PTUV): How It Works
- Underfrequency load shedding (UFLS): how the scheme works
- Automatic service restoration after a loss of voltage
- Fast bus transfer for motor buses: three methods
- Substation interlocking: types, levels and logic
- FLISR: self-healing distribution networks
References
- NERC PRC-010-2, Undervoltage Load Shedding.
- IEEE C37.2, Standard for Electrical Power System Device Function Numbers, Acronyms, and Contact Designations (function 27).
- IEC 61850-7-4, Compatible logical node classes and data object classes (PTUV, PTRC).
- P. Kundur, Power System Stability and Control, McGraw-Hill, chapter on voltage stability.
Frequently asked questions
Why is underfrequency load shedding not enough?
Because frequency is the same across the whole interconnected system and responds to the active power balance, while voltage is local and responds to the reactive power balance. An area can suffer voltage collapse with normal frequency. Undervoltage load shedding acts exactly where the problem is.
How does the scheme tell voltage collapse from a fault-induced sag?
By time and by supervision. A fault sag lasts as long as the protection takes to clear the fault, normally a fraction of a second; the undervoltage that precedes a collapse is sustained for seconds. The scheme also usually requires undervoltage on all three phases and is blocked on a voltage circuit failure or while overcurrent elements detect a fault.
Who defines the stages and settings of a UVLS scheme?
The planner or the grid operator, based on voltage stability studies. In North America, NERC PRC-010 sets the framework for the design, assessment and operation of UVLS programs. Elsewhere, the system operator decides whether a scheme is required and on what criteria.