Electric Power Distribution

An overhead distribution system carries bare (or covered) conductors on poles, using air as insulation. An underground system uses insulated cables laid in trenches, ducts or tunnels below the ground.

Comparison of features

Suitability

• Overhead suits rural and semi-urban areas: low load density, long distances, open space, and limited budget. Examples in Nepal are most 33 kV and 11 kV feeders and LT lines outside city cores.
• Underground suits dense urban areas: high load density, narrow streets, high land and space value, need for safety and appearance, and important consumers needing high reliability. Examples are city centres, airports, industrial estates and heritage sites. The NEA underground distribution projects in Kathmandu Valley (Ratnapark, Maharajgunj and other areas) are examples.
• Hilly terrain, river crossings and flood-prone ground favour overhead lines. Areas with heavy lightning or tree cover favour underground cables or covered conductors.

In practice the choice is economic: underground becomes justified when the load density is high enough that its cost per kW served, plus the value of better reliability, beats the overhead system.

In dense urban areas, underground distribution is often unavoidable for space and safety reasons. When the full costs are counted, it can also be cheaper per unit of load served than overhead lines, while in rural areas the opposite holds.

Why it is a necessity in urban areas:

• Streets are narrow and buildings close, so it is hard to keep pole and conductor clearances from balconies and windows.
• The public safety risk of bare conductors is high.
• Appearance matters (city centres, heritage zones such as Kathmandu Durbar Square).
• Many feeders and transformers must fit into a small space.

Why it can be economical in urban areas:

1. High load density (MW/km²): one cable route serves a large load over a short length. The high cost per km is spread over many kW, so the cost per kW is low. In rural areas the load is small and scattered, so the cost per kW of cable is very high.
2. Short feeder lengths: urban feeders are a few km long, while rural feeders run tens of km. Cable cost grows with length.
3. Overhead is costly in cities too: overhead lines there need short spans, special poles, covered conductors or ABC, tree trimming and more frequent repairs.
4. Reliability value: fewer faults and less outage time mean less lost revenue (kWh not sold). Urban commercial and industrial customers have a high cost of interruption, and loop or ring-main cables restore supply quickly.
5. Lower losses and maintenance: cables have lower reactance, less theft and fewer weather-related faults. These savings over 40 years or more offset the higher first cost.
6. Land value: no pole space or right-of-way is needed on costly urban land.

So for the same total load, an urban underground network gives a lower life-cycle cost per kW than in rural areas. Rural areas, with low density, long distances and cheap open land, remain economical with overhead lines.

In a radial system each feeder starts from the substation and branches out to the loads, so power flows in one direction only. In a loop (ring) system the feeder starts from the substation, passes through the load points and returns to the same or another source. Each load can then be fed from two directions.

 Radial:  S/S ====+=====+=====+==== end
                  L1    L2    L3

 Loop:    S/S ====+=====+=====+====+
           ||     L1    L2    L3   ||
           ++======================++

Loops are often built as open loops: they are run radially, with a normally-open point, and the open point is closed only after a fault is isolated. This gives loop reliability with radial simplicity.

FALSE.

A network (mesh) system is the most reliable distribution system, and a radial system is the least reliable.

• Radial: each consumer has only one supply path. A fault anywhere on the feeder, or on the single distribution transformer, interrupts every consumer beyond that point until it is repaired.
• Network: feeders and secondary mains are interconnected and fed from several transformers and feeders through network protectors. A fault on one feeder or transformer is cleared automatically, and the load is carried by the remaining paths without interruption. A network can withstand a single (and often a double) contingency.

The radial system is preferred for being cheap and simple, not for its reliability.

The two layouts differ mainly in where the distribution transformer is placed and in how much of the system is LT (secondary).

American system

• Long single-phase primary laterals (for example 7.2–14.4 kV) are taken close to every few consumers.
• Many small single-phase pole transformers (10–50 kVA) are used, each feeding 5–10 consumers.
• Very short secondary at 120/240 V.
• Primary is about 70–80% of the network and secondary about 20–30%.

European system

• A three-phase primary (11 kV or 20 kV) runs to fewer, larger three-phase transformers (100–630 kVA).
• Long three-phase 4-wire LT secondary at 400/230 V feeds a whole locality, often 100–500 consumers.
• The secondary forms a large part of the network.

Suitability and choice in Nepal

• Nepal's standard is 400/230 V, 50 Hz, and NEA's network is basically the European type: 33 kV and 11 kV three-phase primaries, three-phase 11/0.4 kV transformers (25–500 kVA), and three-phase 4-wire LT lines. This suits dense urban and semi-urban areas (Kathmandu Valley, Terai towns), where three-phase motors, mills, industries and commercial loads need three-phase supply.
• Its weaknesses in Nepal are long LT lines with high technical losses, low tail-end voltage and electricity theft by hooking. These are major reasons for NEA's high distribution losses.
• For sparse rural and hilly areas with mainly lighting and small domestic loads and long distances, an American-type layout is more suitable. Small single-phase or small three-phase transformers placed close to consumer clusters keep the LT short, cut losses and theft, and improve voltage. NEA's practice of adding more, smaller transformers ("transformer near the load") and its rural electrification schemes move in this direction.

Choice: keep the European layout for urban and industrial areas because of the three-phase demand and high load density. Use an American-type (small transformers, short LT) layout for scattered rural and hilly loads to reduce losses and improve voltage.

A distribution transformer is the step-down transformer that converts the primary distribution voltage (11 kV or 33 kV in Nepal) to the consumer utilization voltage (400/230 V). It supplies the secondary (LT) network.

It is usually pole-mounted or plinth-mounted, rated about 25–500 kVA (for example 11/0.4 kV, 100 kVA, Dyn11). It is connected all the time, so it is designed for low core (no-load) losses and high all-day efficiency.

A load center is the point in a supply area at which the total load can be considered concentrated. It is found as the "centre of gravity" of the loads, with coordinates $\bar x = \sum P_ix_i / \sum P_i$ and $\bar y = \sum P_iy_i / \sum P_i$.

The substation or distribution transformer is placed at or near the load center. This keeps the feeder lengths, voltage drop and $I^2R$ losses to a minimum.

Secondary distribution is the part of the distribution system that operates at the utilization voltage: 400 V three-phase (four-wire) or 230 V single-phase in Nepal. It carries power from the distribution transformer through LT lines and service mains to the consumers' premises.

It is mostly overhead LT lines, ABC or cables, run radially from each transformer.

Primary distribution is the part of the distribution system that operates at medium voltage: 33 kV and 11 kV in Nepal. It carries power from the distribution (grid) substation to the distribution transformers, or directly to large consumers such as industries.

It consists of the primary feeders and laterals, which are usually three-phase, three-wire and mostly radial or open-loop.

The INPS (Integrated Nepal Power System) uses a radial distribution system, of the European type, for both primary and secondary distribution. Only some urban feeders are built as open loops (ring mains) but are still operated radially.

Structure in the INPS

 Grid S/S 132(66)/33/11 kV
     |
 33 kV feeders (radial, long)
     |
 33/11 kV distribution S/S
     |
 11 kV feeders (radial; some urban
     |          feeders open-loop)
 11/0.4 kV transformers (25-500 kVA)
     |
 400/230 V, 3-ph 4-wire LT (radial)
     |
 consumers

• Primary: 33 kV and 11 kV, three-phase, three-wire overhead feeders from grid substations. They are mostly radial with branches (tee-offs). In Kathmandu Valley and other cities, some 11 kV feeders are interconnected with normally-open tie switches or ring main units (open loop). They are operated radially, and the tie is closed only to transfer load after a fault.
• Secondary: 400/230 V three-phase 4-wire LT lines from each 11/0.4 kV transformer, purely radial, mostly overhead (bare conductor or ABC), with underground cables in some urban areas.

Suitability, based on the types of distribution systems

Why radial suits Nepal:

1. The load is scattered and growing in rural and hilly areas, where long single feeders are the only economical option.
2. It has the lowest capital cost per consumer, which matters for NEA's rural electrification and limited investment.
3. Protection and operation are simple, which suits local staff skills and equipment.
4. It is easy to extend as new villages are electrified.

Limitations:

• Low reliability: a fault on the trunk cuts off the whole feeder.
• Poor voltage at the tail ends of long 11 kV and LT lines.
• High losses.

These are being reduced by sectionalizing, open-loop ties in cities, more 33/11 kV substations, shorter LT lines (more transformers near the load), and underground ring-main projects in Kathmandu Valley.

So the INPS uses radial distribution because it is the most economical and practical for Nepal's mostly low-density, spread-out load. Open-loop arrangements are added where load density and reliability needs justify them.

TRUE.

A loop (ring) feeder needs extra conductor to close the ring and extra switches to sectionalize it. Its cost is justified by the load it serves and the reliability it adds.

• Urban area: the load density is high and the load points are close together, so the extra length needed to close the loop is short. The added cost per kW served is small. Urban consumers (commercial, hospitals, offices) also lose a lot from outages, and the loop reduces outage time and losses (current flows from both sides, so the voltage drop is lower). The extra cost is therefore quickly recovered.
• Rural area: the loads are small and scattered over long distances. Closing a loop may need many extra km of line for a few kW of load, so the cost per kW is very high. The value of the energy saved from shorter outages is low. A simple radial feeder is more economical there.

So the loop arrangement is relatively more economical in urban areas than in rural areas.

A radial system feeds the loads from one end only. A loop (ring) system feeds each load from two directions through a closed or normally-open ring. A network (mesh) system interconnects many feeders and transformers on the secondary side, so each load has several supply paths.

 Radial:   S ===+===+===+  (one path)

 Loop:     S ===+===+===+
           |            |   (two paths)
           +============+

 Network:  S1 ==+===+===+== S2
                |   |   |
              ==+===+===+==  (mesh, many paths)

Radial system

Merits

• Simplest and cheapest system, with the lowest initial cost.
• Simple protection (fuses, OC relays) and easy operation.
• Easy to plan and extend.

Demerits

• Low reliability: a fault cuts off all consumers beyond it.
• The far-end consumers get low voltage and suffer voltage fluctuation as the load changes.
• High losses on long feeders.

Loop (ring) system

Merits

• Each load has two supply paths. A faulty section is isolated and supply is restored from the other side, giving good reliability.
• Better voltage regulation and lower losses, because the load current divides between the two paths.
• Maintenance can be done without long outages.

Demerits

• Higher cost (extra conductor, sectionalizing switches, RMUs).
• More complex protection (directional relays) when the loop is operated closed.
• Each section must be sized for the full load when the loop is fed from one side.

Network system

Merits

• Highest reliability: a single feeder or transformer outage causes no interruption.
• Best voltage regulation and lowest losses, since the load is shared among many paths.
• Good flexibility for load growth, and the transformers share load efficiently.

Demerits

• Highest cost: many interconnections, network protectors and more transformers.
• Complex design, protection and fault-current coordination; high fault levels.
• Justified only for very high load density.

TRUE.

In Nepal (NEA's distribution under the INPS), both levels are operated radially:

• Primary distribution (33 kV and 11 kV): feeders leave the grid or distribution substation and branch out to the 11/0.4 kV transformers along one path. Some urban 11 kV feeders in Kathmandu Valley and other cities have tie switches or ring main units forming loops. These are kept normally open (open loop) and the feeders are still operated radially. The tie is closed only to transfer load during faults or maintenance.
• Secondary distribution (400/230 V): each distribution transformer feeds its own LT lines radially to the consumers. There is no interconnection between transformers.

Reasons for radial operation:

1. It has the lowest cost for Nepal's low and scattered load density and long hilly feeders.
2. Protection is simple (drop-out fuses, OC/EF relays, reclosers) and operation is easy.
3. It is easy to extend for rural electrification.

Its main drawbacks are lower reliability, poor tail-end voltage and higher losses. NEA reduces these through sectionalizing, open-loop ties in cities and more substations.