N-1 analysis repeats the load flow once for each element that can go out of service (a line, a transformer, a generator) and checks whether the network stays within its loading and voltage limits. A network meets N-1 when no single outage produces overloads or out-of-band voltages.
Short answer
N-1 analysis repeats the load flow once for each element that can go out of service: a line, a transformer or a generator. In each repetition two things are checked: whether any element ends up above its rating and whether any voltage leaves its band. The network meets N-1 when no single outage produces a violation. The result is a short list of the outages that do, with the affected element and its percentage loading. That list decides where to reinforce, where to redispatch generation or where to accept an agreed load shed.
Why it is done
A load flow with every element in service (the base case, or N-0) can look perfect, and the network can still be one outage away from an overload. Lines trip on faults, are taken out for maintenance or are lost to weather. If the network holds only with everything in service, each of those outages becomes a supply interruption. N-1 measures that fragility before it happens.
How it runs
- Base case. Load flow of the condition under study, for example maximum demand. If the base case already has violations, they are corrected first.
- Contingency list. All the lines and transformers of the study area and, depending on the criterion, the generators. In large networks it is narrowed to those that make sense.
- One outage at a time. The element is removed from the model and the flow is recalculated.
- Limit check. Loading of each element against its emergency rating and voltage of each bus against its band. Emergency limits are usually wider than normal-operation limits; the grid operator or the project sets them.
- Results table. For each outage: the worst element, its percentage loading and whether there is a violation.
- Measures. For each violation an action is proposed and the case is rerun to confirm it.
Network analysis programs have this routine automated: they take the list of elements, disconnect them one by one and store the maximum and minimum loading and voltage across all the cases. The criterion does not change with the program.
DC flow and AC flow
The DC flow assumes 1 per-unit voltages, negligible resistances and small angles. With that, the flow through a line is the angle difference between its ends divided by its reactance, and the system is linear. It is fast and gives good loading values, but it does not see voltages or reactive power. The AC flow solves the full network (with voltages and reactive power) and is the one used for critical outages and to check the voltage band. The usual practice is to screen with DC and confirm with AC.
Worked example: three buses
Example data. 100 MVA base. Bus 1 has the generator and is the reference (angle 0). Buses 2 and 3 are loads of 60 MW and 80 MW. Three lines: L12 (reactance 0.10 pu, rating 150 MW), L13 (0.20 pu, 120 MW) and L23 (0.10 pu, 100 MW). A DC flow is used.
Base case (N-0). The susceptances are 1/x: b12 = 10, b13 = 5, b23 = 10. The equations of buses 2 and 3, with the loads in pu (-0.6 and -0.8), are:
- 20·θ2 − 10·θ3 = −0.6
- −10·θ2 + 15·θ3 = −0.8
The determinant is 20·15 − 10·10 = 200. This gives θ2 = (15·(−0.6) + 10·(−0.8)) / 200 = −0.085 rad and θ3 = (20·(−0.8) + 10·(−0.6)) / 200 = −0.110 rad.
Flows: L12 = (0 − (−0.085))·10 = 0.85 pu = 85 MW (57 % of 150); L13 = 0.110·5 = 0.55 pu = 55 MW (46 % of 120); L23 = (−0.085 + 0.110)·10 = 0.25 pu = 25 MW (25 % of 100). Check: 85 + 55 = 140 MW, which is the total load. The base case passes.
Loss of L12. All the load arrives through L13, and L23 feeds bus 2: L13 = 140 MW and L23 = 60 MW (from bus 3 to bus 2). L13 is at 140 / 120 = 117 %: violation.
Loss of L13. L12 = 140 MW (93 % of 150) and L23 = 80 MW (80 % of 100). It passes.
Loss of L23. L12 = 60 MW and L13 = 80 MW (67 % of 120). It passes.
- Which line goes out of service?
- L12: L13 rises to 140 MW, 117 % of its 120 MW rating, and there is a violation
- L13: L12 is at 93 % and L23 at 80 %, and it passes
- L23: L12 is at 40 % and L13 at 67 %, and it passes
Possible measures. (a) Reinforce L13 to a rating of 150 MW or more; (b) redistribute generation or loads so that, if L12 trips, L13 does not exceed 120 MW (20 MW has to come off bus 3); (c) agree with the customer an automatic shed of 20 MW at bus 3 when L12 trips. Each option is rerun in the model to confirm the violation is gone.
What it delivers and how to read it
Ask for the report to include the base case, the list of contingencies analyzed with their result, the limits used and where they come from, and the proposed measures with the result after applying them. Read the violations first; then check that the contingency list includes the elements you consider critical and that the study ran, at a minimum, maximum and minimum demand.
Checklist
- The base case meets limits before the contingencies are run.
- The outage list includes all the lines and transformers of the area.
- The emergency limits are written down and come from the operator or the project.
- Both loading and voltage are checked, not just one of them.
- At least maximum and minimum demand were run.
- The critical outages from the DC flow were repeated in AC.
- Each violation has a measure and its result after applying it.
Related guides
- Load-flow study: what it delivers and when it is required
- Bus types in load flow: swing, PV and PQ
- Gauss-Seidel and Newton-Raphson in load flow
- Correcting low voltage in a load flow: taps and Q
- Electrical studies with software: what to ask for and review
- Stability of an industrial network with its own generation
References
- Wood, A. J., Wollenberg, B. F. and Sheblé, G. B., Power Generation, Operation and Control, 3rd ed., Wiley, 2013: chapters on security analysis and contingencies.
- pandapower public documentation, section Contingency analysis (describes an N-1 routine over lines and transformers).
Do you need this study?
We run contingency analysis as part of load-flow and connection studies, with the model of your network and a report that lists each outage, its result and the proposed measure. It is engineering design and studies work done remotely, valid for projects in any country. If you need it to connect to a grid, see the connection studies. Contact us.
Frequently asked questions
What does it mean for a network to meet the N-1 criterion?
That with any one of its elements out of service (one at a time) the load flow stays within limits: no line or transformer exceeds its emergency rating and no voltage leaves its permitted band. How many elements and which limits count is set by the grid operator or the project criterion.
Is N-1 the same as a short-circuit study?
No. Short circuit calculates the current of a fault. N-1 assumes an element has already gone out of service, because of a cleared fault or maintenance, and looks at how flows redistribute in the remaining network. They complement each other: the fault causes the outage and N-1 checks the state afterwards.
Is a DC load flow enough for N-1?
It works for a quick first check of loading, because it is linear and needs no iteration. It does not calculate voltages or reactive power. So if the project has voltage limits to meet, the outages that the DC flow marks as critical are repeated with an AC load flow.