PTDF, LODF and DC power flow screening, explained

What a PTDF measures

A power transfer distribution factor (PTDF) is the share of a power transfer between two points that flows on a particular transmission facility, expressed as a percentage of the transfer (Overbye, quoting the NERC definition). A PTDF of 20% on a line means that injecting 100 MW at your point of interconnection and withdrawing it elsewhere adds about 20 MW to that line.

In the linear (DC) model, a PTDF does not depend on the size of the transfer, and its magnitude is at most 1 (Kekatos). That is what makes it cheap to compute across thousands of facilities.

What an LODF adds

A line outage distribution factor (LODF) is the portion of a line's pre-outage flow that moves onto another line when the first line is taken out of service (Overbye). With it, the post-outage flow on line ℓ after losing line k is the pre-outage flow on ℓ plus LODF(ℓ,k) times the pre-outage flow on k (Kekatos).

Combining the two gives the outage transfer distribution factor (OTDF): the PTDF of a transfer on line ℓ with line k out of service. Contingency screening in interconnection studies relies on these combinations.

How grid operators use distribution-factor thresholds

In cluster studies, a distribution factor threshold decides whether a request is counted as contributing to an overload, and therefore whether it can be assigned a share of the upgrade.

SPP also allocates upgrade costs in proportion to each request's distribution factor multiplied by its megawatts (SPP). A small change in a distribution factor can therefore move a request across a threshold, or change its share of a large upgrade.

What a DC screen cannot tell you

DC power flow assumes lossless lines, voltage magnitudes that are known and constant, reactive power that can be ignored, and small angle differences. Resistive losses are its largest source of error, and accuracy falls as the ratio of resistance to reactance rises (Simpson-Porco). Distribution factors also do not work well for reactive power (Overbye).

• A DC screen can show real-power (MW) sensitivities and likely thermal overloads.
• It cannot show voltage, reactive power, stability or short-circuit problems, which operators study separately.
• Its answer is only as current as the network model, the dispatch assumptions and the queue it was run against.

Use a DC screen to rank candidate sites and to ask better questions, then record its model version, date and limits next to the result. Formal studies remain authoritative. For the site comparison itself, see how to choose a point of interconnection.