Description
In edisgo/equipment/equipment-parameters_LV_transformers.csv the copper loss P_k
of the 1000 kVA standard transformer is clearly wrong — most likely a misplaced
decimal point. As a result the resistance of this transformer is modelled about a
factor of 10 too low in the network model.
Affected file
edisgo/equipment/equipment-parameters_LV_transformers.csv
name,S_nom,u_kr,P_k
#,MVA,%,MW
100 kVA,0.1,4,0.00175
160 kVA,0.16,4,0.00235
250 kVA,0.25,4,0.00325
400 kVA,0.4,4,0.0046
630 kVA,0.63,4,0.0065
800 kVA,0.8,6,0.0084
1000 kVA,1.0,6,0.00105 <-- 1.05 kW, should be ~10.5 kW (0.0105)
Why this is a bug
P_k (short-circuit / copper losses in MW) feeds into the per-unit resistance of the
transformer:
r_pu = P_k / S_nom — see edisgo/network/topology.py:213
(_load_equipment_data)
and from there into x_pu = sqrt((u_kr/100)^2 - r_pu^2).
Sanity check across the whole table:
| Type |
P_k [kW] |
r_pu = P_k/S_nom |
| 100 kVA |
1.75 |
0.0175 |
| 160 kVA |
2.35 |
0.0147 |
| 250 kVA |
3.25 |
0.0130 |
| 400 kVA |
4.6 |
0.0115 |
| 630 kVA |
6.5 |
0.0103 |
| 800 kVA |
8.4 |
0.0105 |
| 1000 kVA |
1.05 ❌ |
0.00105 ❌ |
The P_k series increases monotonically (1.75 → 8.4 kW); the 1000 kVA value (1.05 kW)
even falls below the 100 kVA transformer and pulls r_pu down by a factor of 10.
This is not physically plausible.
Standard distribution transformers per EN 50464-1 have load losses of roughly
10.5 kW at 1000 kVA. The other rows of the table match those reference values
(e.g. 100 kVA = 1750 W, 630 kVA = 6500 W).
Expected behaviour
P_k for 1000 kVA should be 0.0105 MW (10.5 kW). This gives r_pu = 0.0105, which
lines up seamlessly with the 800 kVA value (0.0105).
Suggested fix
-1000 kVA,1.0,6,0.00105
+1000 kVA,1.0,6,0.0105
Impact
When a 1000 kVA distribution transformer is installed during grid reinforcement
(standard equipment / selection in reinforce_measures), its resistance is modelled
too low. Consequences:
- The copper losses of this transformer are underestimated in the power flow.
- The resistive voltage drop across the transformer is underestimated → potentially
too optimistic voltage results and, in turn, possibly an understated reinforcement
need whenever 1000 kVA transformers are involved.
x_pu changes only marginally (≈ 0.0600 → 0.0591) and stays real — so no crash,
just silently wrong physics.
Description
In
edisgo/equipment/equipment-parameters_LV_transformers.csvthe copper lossP_kof the 1000 kVA standard transformer is clearly wrong — most likely a misplaced
decimal point. As a result the resistance of this transformer is modelled about a
factor of 10 too low in the network model.
Affected file
edisgo/equipment/equipment-parameters_LV_transformers.csvWhy this is a bug
P_k(short-circuit / copper losses in MW) feeds into the per-unit resistance of thetransformer:
r_pu = P_k / S_nom— seeedisgo/network/topology.py:213(
_load_equipment_data)and from there into
x_pu = sqrt((u_kr/100)^2 - r_pu^2).Sanity check across the whole table:
The P_k series increases monotonically (1.75 → 8.4 kW); the 1000 kVA value (1.05 kW)
even falls below the 100 kVA transformer and pulls
r_pudown by a factor of 10.This is not physically plausible.
Standard distribution transformers per EN 50464-1 have load losses of roughly
10.5 kW at 1000 kVA. The other rows of the table match those reference values
(e.g. 100 kVA = 1750 W, 630 kVA = 6500 W).
Expected behaviour
P_kfor 1000 kVA should be0.0105MW (10.5 kW). This givesr_pu = 0.0105, whichlines up seamlessly with the 800 kVA value (0.0105).
Suggested fix
Impact
When a 1000 kVA distribution transformer is installed during grid reinforcement
(standard equipment / selection in
reinforce_measures), its resistance is modelledtoo low. Consequences:
too optimistic voltage results and, in turn, possibly an understated reinforcement
need whenever 1000 kVA transformers are involved.
x_puchanges only marginally (≈ 0.0600 → 0.0591) and stays real — so no crash,just silently wrong physics.