Automation

Ethernet-APL network sizing

In an Ethernet-APL network with locally powered field switches, what gets sized is each spur — length by cable category, compatibility between the power classes of port and device, and the voltage reaching the device with the cable hot —, the ports and consumption of each field switch, and the cyclic network load. This page shows the calculations and the limits of the Ethernet-APL Engineering Guideline.

When to use

When designing the instrument network for PROFINET, EtherNet/IP or HART-IP over Ethernet-APL, when choosing the spur cable, when sizing the auxiliary supply of the field switch cabinets and when preparing the 2-WISE descriptive document of a hazardous-area installation.

What gets sized in an Ethernet-APL network

In the structure sold today, the control network arrives in Industrial Ethernet — copper or fibre — at field switches installed in a cabinet or junction box and powered by a local supply. From each field switch runs one APL spur per device — flow meter, differential pressure transmitter or radar level transmitter: a single pair carrying data at 10 Mbit/s and power, the same two-wire idea as a PROFIBUS PA segment but with Ethernet all the way to the instrument. Sizing has three parts: the spurs, the field switches and the network load.

The spur: length, class and voltage

The maximum length depends on the cable category — up to 200 m on new type A cable. But the spur also has to deliver power: the port guarantees a minimum voltage at the maximum current of the class, the cable drops I × R, and the minimum operating voltage must remain at the device. For class A that margin is only 0.6 V; at 55.56 mA about 10.6 Ω of loop fit. That is why, with thin and hot cable, the spur limit is voltage, not the 200 m.

The power class of the port must be compatible with that of the device: a C port feeds A and C devices, an A port feeds only A devices.

The field switch: ports and power

Each field switch has a fixed number of ports, and it pays to leave spares. The consumption from the auxiliary supply is a fixed part plus a part per used port, larger on class C ports. Adding the switches of a cabinet gives the power of the 24 V supply; the feeder from that supply is a DC cable sizing problem in its own right, and a supply that must ride through outages is backed by a battery bank. On the control-room side, the Industrial Ethernet switches and fibre patch panels that feed the field switches go into a 19” rack.

Hazardous area

With 2-WISE the spur is intrinsically safe without entity parameter calculation, as long as source, load and auxiliaries are 2-WISE and the cable is within the resistance, inductance and capacitance ranges. The protection level of the segment is that of the least demanding port, and the required documentation is a descriptive document — the guideline gives a model in Table 10-1.

What this tool does not cover

The powered APL trunk, in which the trunk itself powers the field switches, is left out: the Engineering Guideline v1.14 states that no calculation scheme is published for it and refers to the vendor.

Formulas and fundamentals

Voltage at the device U = U_PS(min) − I_class × R_loop,20 × [1 + α20 (θ − 20)] × L ≥ U_PL(min)

The source port guarantees U_PS(min) at the maximum current of the load class; the cable drops I × R and U_PL(min) must remain at the device. α20 = 3.93×10⁻³/K, θ is the maximum cable temperature.

Maximum spur by voltage L_max = [U_PS(min) − U_PL(min)] / [I_class × R_loop(θ)]

The length at which the margin reaches zero. The final maximum spur is the smaller of this value and the cable category limit (50 to 200 m).

Field switch consumption P = P_fixed + P_port × used ports

Power drawn from the auxiliary supply. In the FieldConnex rail field switch datasheets the fixed part is close to 10 W and each used class A port costs about 1 W.

Network load load = n × (payload + 42 bytes) × 8 / (cycle × data rate)

Eq. 10-1/10-6 of the guideline. With the minimum payload of 46 bytes each frame has 88 bytes. Below 20 %, nothing to do; 20 to 50 %, check; above 50 %, reduce.

Standards & methods

  • Ethernet-APL Engineering Guideline v1.14 (FieldComm Group, ODVA, OPC Foundation, PI)
  • Ethernet-APL Port Profile Specification — power classes
  • IEEE 802.3cg (10BASE-T1L)
  • IEC TS 60079-47 — 2-WISE
  • IEC 60228 — conductor resistance

Typical reference values

Quantity Typical range Note
Class A — U_PS(min) / I / U_PL(min) 9.6 V / 55.56 mA / 9.0 V —
Class C — U_PS(min) / I / U_PL(min) 11.61 V / 95 mA / 10.6 V —
Allowed source → load A → A · C → A and C —
Maximum spur by cable category I 50 m · II 100 m · III 150 m · IV 200 m —
Auxiliaries / insertions on the spur up to 2 / up to 4 —
2-WISE cable loop 15–150 Ω/km · 0.4–1 mH/km · 45–200 nF/km spur up to 200 m, up to 2 auxiliaries
Fieldbus type A cable (0.8 mm²) loop ≈ 44 Ω/km manufacturer datasheet

Worked example

Long spur in thin cable

Inputs

Cable
0.75 mm² (IEC 60228)
Maximum cable temperature
55 °C
Port and device
class A —
Spur
200 m

Results

Loop resistance at 55 °C
11.15 Ω
Voltage at the device
8.98 V (minimum 9.0 V)
Maximum spur with 0.75 mm²
194 m
Cross-section that complies
1 mm²

The 200 m spur is within the category IV limit, but the voltage does not close: 8.98 V arrive for a 9.0 V minimum. There are three ways out: shorten to at most 193.7 m, move to 1.0 mm² (the 200 m then deliver 9.14 V), or use a cable whose real loop resistance, per datasheet, is below the IEC 60228 maximum.

Run your own calculation Create a free account and get this for your own inputs in seconds.

Common mistakes

  • Looking only at length. 200 m is the category IV limit, but voltage may fail first: with 0.75 mm² at 55 °C a class A spur stops at 193.7 m.
  • Calculating with a cold cable. Resistance grows about 14 % between 20 °C and 55 °C; the spur that passes in the office fails on the sunny pipe rack.
  • Connecting a class C device to a class A port. The A port does not guarantee the voltage or current it needs; the combination is forbidden by Table 4-5.
  • Sizing the cabinet supply by the no-load consumption of the switch. Each used port adds 1 to 2 W; with 24 ports the difference exceeds 20 W.
  • In a hazardous area, forgetting that the segment level is that of the least demanding port. An Ex ia device on an Ex ic port gives an Ex ic segment.

Frequently asked questions

What is the maximum length of an Ethernet-APL spur?

Up to 200 m in category IV cable, less in lower categories (50, 100 and 150 m). But the limit may be voltage rather than length: with thin, hot cable the spur stops short of 200 m.

What is the difference between classes A and C?

Class C delivers more power to the device (11.61 V and 95 mA guaranteed at the port, against 9.6 V and 55.56 mA for class A). A C port feeds A and C devices; an A port feeds only A devices.

What about the powered APL trunk?

The Engineering Guideline v1.14 states that no calculation scheme is published for the powered trunk and refers to the vendor. This tool covers the structure sold today: Industrial Ethernet up to locally powered field switches.

Is APL network load usually a problem?

Rarely in process applications. Per the guideline example, 50 sensors at 200 ms take 1.76 % of a 10 Mbit/s link. What matters is a very short cycle or a large payload.

What is 2-WISE?

The intrinsic safety concept of IEC TS 60079-47 for Ethernet-APL spurs. With 2-WISE source, load and auxiliaries and cable within the R, L and C ranges, the spur is intrinsically safe without entity parameter calculation — the descriptive document is enough.

Glossary

Spur
Point-to-point link between a field switch port and a field device, up to 200 m.
Field switch
Field switch converting Industrial Ethernet into APL ports and powering the devices through the spurs.
Power class
Voltage and current range an APL port guarantees (source) or requires (load), such as A and C.
2-WISE
Intrinsic safety concept of IEC TS 60079-47 for Ethernet-APL spurs.
10BASE-T1L
IEEE 802.3cg physical layer — 10 Mbit/s full duplex over a single pair, up to 1,000 m.