Two medium-voltage (MV) networks can use the same cables, the same transformers and the same switchgear ratings, yet perform very differently when a cable fails. In one, every customer downstream of the fault waits for a repair crew. In another, supply is restored within the hour by moving a single switch. In a third, the customers may not notice the fault at all. The difference is the network configuration: how feeders, switching points and substations are connected, and how they are normally operated.
This article answers a practical question: how do radial, open-ring and meshed MV distribution networks operate, and how does the chosen arrangement shape continuity of supply, protection and the equipment you need to specify?
Key takeaways
- The network configuration determines what happens after a fault: how many customers lose supply, for how long, and which devices must detect and clear it.
- Radial networks are the simplest and least expensive, but a fault on a feeder interrupts everything downstream of the protective device that clears it.
- Open-ring networks are built as loops but operated radially, with a normally-open point (NOP). Supply to healthy sections can be restored by switching, after a short interruption.
- Closed-ring and meshed (interconnected) networks keep several supply paths in service at once. They can keep healthy sections supplied through a fault, which typically relies on circuit-breakers at the ends of the links that must be cleared automatically and on directional or unit protection; the exact arrangement depends on the particular design.
- Configuration drives equipment choice: radial and open-ring secondary substations typically use ring main units (RMUs) or compact / package substations; meshed networks typically use circuit-breaker switchgear panels with protection relays at their switching nodes.
- Every arrangement needs the switchgear at each location to be rated for the fault level, current and switching duties at that point — and those values can change when the network is reconfigured.
Why network configuration matters
An MV distribution network carries power from a primary substation — where a transmission or sub-transmission supply is transformed down to MV — to secondary substations, where MV is transformed to low voltage (LV) close to the loads, and to large consumers supplied directly at MV.
Network designers balance four things:
- Continuity of supply — how often customers are interrupted, and for how long.
- Cost — cable routes, number of switching devices, protection and automation.
- Operational complexity — switching procedures, protection settings and the skills needed to run the network.
- Flexibility for growth — how easily new loads and new sources (for example solar PV or battery storage) can be added.
No configuration wins on all four. The choice is usually set by the network operator or the project’s design consultant, but everyone involved in specifying and procuring equipment needs to understand it, because it fixes the number and type of switching functions, the protection philosophy and the ratings each piece of switchgear must meet.
Building blocks of an MV distribution network
Before comparing configurations, it helps to define the elements that appear in all of them.
Voltage levels
In IEC terms, the voltage by which a network is identified is its nominal system voltage; equipment is rated for the highest voltage for equipment. IEC 60038 lists standard values, for example 11 kV, 22 kV and 33 kV (and 10, 20 and 30 kV) paired with equipment voltages of 12, 24 and 36 kV. Nominal voltages used in practice vary by country and utility. The relationship between nominal voltage and switchgear rated voltage is explained in Specifying MV Switchgear: Key Ratings Explained.
Source, feeders and substations
- Source busbar — the MV busbar in the primary substation (or in a customer’s main MV switchroom), fed from one or more HV/MV or MV/MV transformers.
- Feeder circuit-breaker — the circuit-breaker at the head of each outgoing feeder, with a protection relay that detects faults and trips the breaker.
- Feeder — the cable or overhead line route that carries power to the substations it serves.
- Secondary substation — where the feeder connects to an MV/LV transformer, often through an RMU or as a complete compact / package substation.
- Switching points — load-break switches, disconnectors or circuit-breakers that allow sections of the network to be isolated, earthed and reconfigured.
Normally-open point and n-1
- A normally-open point (NOP) is a switch that is deliberately kept open in normal operation so that a looped network is operated radially. It is typically closed only to transfer load or restore supply; where NOPs are located and how they are operated is set by the network design and the operator’s procedures.
- The n-1 principle describes a network that can lose any one element (a cable, a transformer, a source) and still supply its load, either automatically or after switching. How quickly that happens depends on the configuration.
Radial, open-ring and meshed: the distinctions used here
- A radial network has a single supply path to each substation by construction; there is no alternative route to switch to.
- An open-ring network is built as a loop, so each substation has two possible supply paths, but it is operated radially: a normally-open point keeps only one path energised at a time, and the second path is brought in by switching.
- A closed-ring or meshed (interconnected) network keeps more than one supply path energised at the same time. A closed ring is the simplest meshed form; a fully meshed network interconnects several sources and nodes.
Terms vary by region and utility (“ring main”, “open loop”, “interconnected”); this article uses them in the sense above.
Radial networks
How a radial network operates
In a radial network, each feeder leaves the source busbar through its own circuit-breaker and supplies its substations along a single path. Power flows in one direction only, from the source towards the end of the feeder. There is no alternative path.
Continuity of supply
A fault anywhere on the feeder is cleared by the feeder circuit-breaker (or by a downstream protective device, where one is fitted). Every substation downstream of the device that clears the fault loses supply. After the fault has been located, the faulted section can be isolated by opening the switches either side of it, and the substations upstream of the fault can be re-energised by reclosing the feeder breaker. Substations downstream of the fault remain without supply until the cable is repaired, unless a temporary supply (for example a mobile generator) is connected.
Switching and protection
Radial protection is comparatively simple, because fault current can only flow one way:
- Time-graded overcurrent and earth-fault protection at the feeder circuit-breaker, coordinated with upstream protection at the source.
- Fuses or switch-fuse combinations protecting individual transformers at secondary substations, coordinated so that a transformer fault blows its own fuse before the feeder breaker trips.
- Load-break switches at substations for isolation and switching; they are not expected to interrupt fault current.
Fault level is highest near the source and falls along the feeder as cable impedance increases.
Typical use and equipment
Radial networks are common for rural and overhead networks, small industrial sites, single-transformer customer connections and the first stage of a development that will later be extended. Secondary substations are often a compact / package substation or an RMU with a transformer protection function. A radial connection can be prepared for later conversion to a ring by specifying an RMU with a spare cable way from the start.
Open-ring networks
How an open-ring network operates
An open ring (also called an open loop or ring main) is built as a loop: a feeder leaves one circuit-breaker at the source, passes through a series of secondary substations and returns to a second circuit-breaker, at the same source or at a different one. In normal operation one switch in the loop — the normally-open point — is kept open. The ring therefore runs as two radial feeders, each supplying the substations on its side of the NOP.
Each secondary substation on the ring usually has an RMU with two ring switches (one for each direction around the ring) and a transformer protection function (a switch-fuse combination or a circuit-breaker with relay).
What happens after a fault
The value of an open ring is that every substation can be supplied from either end. Consider a cable fault between two substations on the side fed by feeder A. The sequence below is an illustrative example; actual restoration procedures, and the order of the steps, are defined by each network operator:
- The feeder A circuit-breaker trips. All substations on the feeder A side of the NOP lose supply; the feeder B side is unaffected.
- Fault passage indicators (where fitted) on the RMUs show which cable section carried the fault current.
- The ring switches on each side of the faulted cable section are opened, and the section is earthed for repair.
- The feeder A circuit-breaker is reclosed, restoring the substations between the source and the fault.
- The NOP is closed, so feeder B now supplies the substations between the fault and the old NOP.
All substations are back on supply; only the faulted cable section remains out of service. The interruption lasts as long as it takes to locate the fault and carry out the switching — from minutes with remote-controlled RMUs and automation, to longer where switching is manual. This is an example of n-1 restoration after switching, not uninterrupted supply.
Protection and rating implications
- Protection remains largely radial: feeder breakers at the two ring heads, with fuse or breaker protection of each transformer.
- Because the NOP can move, every ring cable and every RMU ring switch must be rated for the load and fault level that occurs when the ring is supplied entirely from one end. A ring designed for half the load on each side may be overloaded when one feeder has to carry everything.
- Protection settings at each ring head must remain valid for the longest feeder configuration expected after switching.
- For example, where the two ring ends are fed from different sources, closing the NOP while both ends are live may momentarily parallel the sources. Whether this is permitted, and under what conditions, depends on the network operator’s design and procedures; it can affect switching procedures and, possibly, protection.
Typical use and equipment
Open rings are the most common arrangement for urban and suburban cable networks, commercial developments, campuses and industrial estates. They make extensive use of non-extendible or extendible RMUs and compact / package substations with integrated RMUs. Remote control, motorised switches and fault passage indicators can be added to shorten restoration time.
Closed-ring and meshed (interconnected) networks
How they operate
In a closed ring, the loop is operated with all switching points closed, so every substation is supplied from both directions at once. A meshed or interconnected network goes further: several sources and several paths are connected together, so power can reach a node by more than one route at all times.
Because power can flow in either direction through each link, a fault fed from two sides cannot be cleared by a single device. The faulted section must be disconnected at both ends. Where a link must be cleared automatically without interrupting the rest of the network, as in the example below, this typically means a circuit-breaker at each end of that link, each controlled by protection that can tell which section is faulted. The actual switching and protection arrangement depends on the particular design: some interconnected networks, for example, place breakers only at selected points and accept switching-based restoration elsewhere.
Continuity of supply
In a design like the one illustrated, when protection works as intended, only the circuit-breakers at the two ends of the faulted link open. All substations continue to be supplied through the remaining paths, typically without an interruption. This is n-1 security without switching, and it is the main reason meshed arrangements are used where continuity is critical.
Protection requirements
Simple time-graded overcurrent protection is not sufficient on its own, because fault current flows from several directions. Typical approaches include:
- Directional overcurrent and directional earth-fault protection, which trips only for current flowing into the protected section. These functions need a voltage (polarising) reference as well as current. The reference commonly comes from voltage transformers, but depending on the design it may be taken from busbar VTs shared by several feeders or from voltage sensors integrated in the switchgear, and some directional earth-fault schemes can use current polarisation instead. A dedicated VT for every feeder is therefore not always required.
- Unit protection, such as line differential protection, which compares current at both ends of a link and trips both ends for an internal fault. This needs a communication channel between the two ends.
- Fault levels: connecting several sources in parallel can increase the fault current available at each node compared with a radial arrangement. The actual result depends on the network configuration, the impedance of the links and the characteristics of the sources (inverter-based generation, for example, contributes limited fault current). Switchgear short-circuit ratings must be checked for the meshed condition.
Typical use and equipment
Meshed and closed-ring arrangements are found in large industrial plants with on-site generation, city-centre networks with high load density, critical infrastructure (data centres, hospitals, airports) and MV collection or distribution networks that must stay in service during a single fault. Each node is usually a switchboard of circuit-breaker panels with protection relays, current transformers, voltage measurement where the protection requires it, and auxiliary supplies, rather than an RMU.
Comparison of configurations
| Criterion | Radial | Open ring | Closed ring / meshed |
|---|---|---|---|
| Continuity after a feeder fault | Downstream supply lost until repair | Short interruption; restored by switching | Healthy sections normally stay supplied |
| Relative capital cost | Lowest | Moderate (extra cable to close the loop, RMUs with two ring ways) | Highest (typically more circuit-breakers, protection and communications) |
| Operational complexity | Low | Moderate (NOP management, switching procedures) | High (protection coordination, settings management) |
| Typical switching devices at secondary substations | Load-break switch, switch-fuse | RMU: two ring switches plus transformer protection | Typically circuit-breaker panels with relays |
| Protection implication | Time-graded overcurrent and earth fault; fuses | Largely radial protection, valid for either feed direction | Directional and/or unit (differential) protection; voltage reference and/or communications as the scheme requires |
| Fault level | Highest at source, falls along feeder | As radial, but must be checked for the worst switching state | Can be higher where sources operate in parallel, depending on configuration and source characteristics |
| Typical use | Rural and overhead feeders, small sites, first project phase | Urban cable networks, commercial and industrial estates | Critical sites, dense city centres, plants with on-site generation |
How network configuration affects continuity, cost, complexity, switching devices, protection and fault level.
These are general tendencies, not fixed rules: utility practice, cable lengths, automation and the number of sources all shift the balance.
How configuration drives equipment selection
Ring main units
An RMU is a compact switchgear unit combining cable switching functions and a transformer protection function in one enclosure. Depending on the design, its switching functions may be gas-insulated, air-insulated or use solid insulation. It is the natural fit for open-ring secondary substations: two load-break ring switches allow the ring to pass through, and the transformer function protects the local MV/LV transformer. In radial networks an RMU may be used with one cable way, or with a spare way for future ring conversion. RMU functions and configurations are covered in detail in a later Engineering Insights article.
MV switchgear panels
Metal-enclosed switchgear panels with circuit-breakers and protection relays are used:
- at primary substations and main MV switchrooms, as feeder breakers at the head of radial feeders and ring ends;
- at the nodes of a closed-ring or meshed network, where the design requires a breaker at each end of a link;
- at large customer substations where the utility or design requires circuit-breaker protection for incomers or transformer feeders.
Their ratings — rated voltage, normal current, short-time and peak withstand current, breaking capacity and internal arc classification — are explained in Specifying MV Switchgear: Key Ratings Explained.
Compact / package substations
A compact / package substation combines MV switchgear (usually an RMU), a distribution transformer and an LV board in one factory-assembled enclosure. It sits at the secondary substation position in radial and open-ring networks, and its MV switchgear configuration should match the network: one or two cable ways for radial service, two ring ways for open-ring service. See Compact / Package Substations: How the Four Components Fit Together.
Ratings that depend on the configuration
| Rating | Why the configuration matters |
|---|---|
| Busbar and ring-way normal current | In open rings, sized for the whole ring load supplied from one end |
| Short-time and peak withstand current | Can be highest in meshed networks with parallel sources, depending on configuration and source characteristics; check every switching state |
| Breaking duty | Assigned to circuit-breakers or fuses; load-break switches are not designed to break fault current |
| Number and type of functional units | Radial: fewer ways; open ring: two ring ways plus transformer; meshed: typically a breaker at each link end, as the design requires |
| Protection and instrument transformers | Directional schemes need a voltage reference (commonly VTs or voltage sensors); both need CT classes suited to protection; unit protection also needs communication links |
Practical selection considerations
When assessing which configuration a project needs, or which equipment fits an existing network, work through these questions:
- What does the network operator require? Where the project connects to a public network, the utility usually defines the connection arrangement, switchgear type and protection. Treat this as the starting point.
- What does an interruption cost? Continuous processes, hospitals and data centres justify higher investment in redundancy than a warehouse or a rural village.
- How many sources are there? One source limits the options; two or more sources, including on-site generation, enable rings and meshes but can raise fault levels, depending on the configuration and the type of source.
- How will faults be located and restored? Manual switching, remote control, fault passage indicators or full automation each change restoration time and equipment content.
- What will the network look like in ten years? A radial connection with a spare RMU way, or a switchboard with spare panels, is cheaper to extend than a replacement.
- What is the fault level in every operating state? Check the highest prospective fault current at each switchgear location, including future reinforcement and generation.
Relevant standards
Network configuration itself is a design and operating decision, usually set by the network operator or the project designer rather than by an equipment standard, although national grid codes or utility standards may prescribe connection arrangements. The equipment placed in each configuration is specified primarily against IEC standards such as these. Confirm the applicable edition for each project, together with any national or utility requirements that modify it.
| Standard | What it covers | Why it matters |
|---|---|---|
| IEC 60038 — IEC standard voltages | Standard nominal system voltages and the associated highest voltages for equipment above 1 kV | Links the nominal voltage of the network (for example 11, 22 or 33 kV) to the rated voltage of the equipment to specify (12, 24 or 36 kV) |
| IEC 62271-1 — Common specifications for switchgear and controlgear | Rated characteristics, standard values and common test requirements shared by MV switching devices | Provides the common rating basis (rated voltage, normal current, short-time and peak withstand currents) used to compare every switching device in the network |
| IEC 62271-200 — AC metal-enclosed switchgear, 1–52 kV | Switchgear panels and RMUs assessed as complete assemblies | Sets the assembly-level properties to specify for switchboards and RMUs at feeder heads and secondary substations, such as internal arc classification |
| IEC 62271-100 / -103 / -105 — Circuit-breakers / switches / switch-fuse combinations | Device-level ratings and switching duties for circuit-breakers, load-break switches and switch-fuse combinations | Confirms that each device suits its position: fault breaking at feeder heads, load switching at ring positions and fused protection of transformers |
| IEC 62271-202 — Prefabricated (HV/LV) substations | Factory-assembled substations combining MV switchgear, transformer and LV equipment in one enclosure | Applies where compact / package substations are used at secondary substation positions |
| IEC 60255 series — Measuring relays and protection equipment | Requirements for protection relays, with individual parts covering specific protection functions | Covers the overcurrent, directional and differential protection functions on which each configuration depends |
| IEC 60909-0 — Short-circuit currents in three-phase AC systems | Calculation method for prospective short-circuit currents | Fault levels change with the switching state of a ring or meshed network; the calculated values set the short-circuit ratings that the equipment must meet |
Specification and procurement considerations
The equipment supplier needs to know where each item sits in the network. A clear specification states:
- the network configuration at the installation point (radial, open ring, closed ring, meshed) and whether the location can be a NOP;
- rated voltage and insulation level, aligned with the network nominal voltage;
- prospective short-circuit current and duration at the installation point, for the worst switching state;
- rated normal current of busbars and of each functional unit, including the one-end-feeding case for rings;
- the functional units required (for example two ring switches plus a transformer switch-fuse, or incomer, bus section and feeder breakers);
- protection philosophy — non-directional, directional or unit protection — and the CT, VT and communication requirements that follow;
- neutral earthing method of the MV network, which affects earth-fault protection and equipment insulation;
- remote control, motorisation and fault passage indication, if required;
- utility approvals — some operators accept only approved equipment types;
- extendibility and spare ways for future connections.
AdepaTrade sources MV switchgear, RMUs and compact / package substations to project specifications; it does not design networks or carry out installation. On request, AdepaTrade can arrange on-site engineering supervision and commissioning support under a separate scope.
Common mistakes
- Specifying ring switches and ring cables for half the ring load, forgetting the one-end-feeding case after the NOP moves.
- Using the radial fault level for switchgear in a network that will later be meshed or connected to on-site generation.
- Expecting load-break switches to clear faults in a closed ring.
- Applying non-directional overcurrent protection where fault current can flow from two directions.
- Omitting the voltage reference (VTs or sensors) or the communication links that the chosen directional or differential scheme needs.
- Choosing a non-extendible RMU where the network is expected to grow.
- Leaving the network configuration out of the RFQ, so the supplier cannot confirm functional units and ratings.
Conclusion
Radial, open-ring and meshed networks are not simply drawing styles; they define how a network behaves when something fails. Radial networks are simple and economical but expose downstream customers to every feeder fault. Open rings add a second supply path that can be brought in by switching, which is why RMUs and compact / package substations dominate urban cable networks. Closed-ring and meshed arrangements keep supply through a fault, typically at the cost of more circuit-breakers, more complex protection and potentially higher fault levels. Knowing the configuration at each installation point is the first step to specifying switchgear with the right functions, ratings and protection.
Related AdepaTrade solutions
AdepaTrade sources MV switchgear, ring main units and protection equipment up to 36 kV, as well as compact / package substations, from manufacturers in Türkiye and worldwide, matched to your network configuration and specification. Medium Voltage Distribution & Protection · Compact / Package Substations
Have a technical question? Talk to our team.