Overlaying DC Transmission On an AC System

, , , and

One of the ideas proposed for developing a transmission super highway is to construct a number of DC lines to overlay the existing AC system. The motivation comes from DC’s inherent capability to overcome voltage and stability issues for transfers over long distances, making the network electrically “smaller.” (There is also an economic motivation that is important to recognize, but which we will not treat in this article.) How would such a massive development change the nature of a power system? What issues do system planners have to consider? In a more global view, what would investors and regulatory agencies need to take into account in funding and approving this type of project?

A recent system reliability impact study (SRIS) of a proposed HVDC line from Albany, NY to New York City offers some key insights. Two 1000-MW HVDC bipoles are proposed.  The DC lines would bring ostensibly cheaper upstate energy to the New York City load area, characterized by high locational energy prices. The bipoles would overlay an existing 345 and 230 kV network, and specifically impact several key AC interfaces in the New York grid: Upstate New York – ConEd (UPNY-ConEd), Central-East, Total East, Upstate New York – Southeast New York (UPNY-SENY) and New York City Cables (NYC Cable). Using New York ISO standards for conducting an SRIS, the study identified the following changes in transfer limits:

Interface Condition MW Incr
Central East Summer Peak Thermal Emergency 685
Central East Summer Peak Thermal Normal 730
Central East Winter Peak Thermal Normal 161
Central East Summer Peak Voltage Transfer 487
Total East Summer Peak Thermal Emergency 730
Total East Summer Peak Thermal Normal 1006
Total East Winter Peak Thermal Normal 98
Total East Summer Peak Voltage Transfer 744
Average 580
UPNY-ConEd Closed Summer Peak Thermal Emergency 2648
UPNY-ConEd Closed Summer Peak Thermal Normal 2522
UPNY-ConEd Closed Winter Peak Thermal Normal 2147
UPNY-ConEd Closed Summer Peak Voltage Transfer 1392
UPNY-ConEd Open Summer Peak Thermal Emergency 2648
UPNY-ConEd Open Summer Peak Thermal Normal 2523
UPNY-ConEd Open Winter Peak Thermal Normal 2147
UPNY-ConEd Open Summer Peak Voltage Transfer 1887
UPNY-SENY Closed Summer Peak Thermal Emergency 2157
UPNY-SENY Closed Summer Peak Thermal Normal 2605
UPNY-SENY Closed Winter Peak Thermal Normal 1801
UPNY-SENY Open Summer Peak Thermal Emergency 2158
UPNY-SENY Open Summer Peak Thermal Normal 2606
UPNY-SENY Open Winter Peak Thermal Normal 2101
Average 2239
NYC Cable Summer Peak Thermal Emergency 1625
NYC Cable Summer Peak Thermal Normal 1625
NYC Cable Winter Peak Thermal Normal 1409
NYC Cable Summer Peak Voltage Transfer 1450
Average 1527

On the average, the increase in transfer capacity on the Central-East and Total East interfaces is 29% of the project’s 2000-MW capacity. Of particular note is the voltage-constrained Central East interface in which capacity increases by 487 MW at summer peak (obtained using P-V curves). The location of the project’s sending terminals are just east of these interfaces, and so the project does not, in fact, offer a new parallel path for these particular interfaces. However, the project redistributes power flows in a way that a previously limiting constraint is relieved, and this accounts for the incremental capacity. Also, since the DC terminals require reactive compensation, the voltage limit would actually decrease, if it were not for the resolution of a limiting contingency.

Across the main interfaces from upstate NY to NYC (UPNY-Coned and UPNY-SENY), there is an average increase in transfer capacity of 112% of the bipoles’ capacity.  Since these interfaces are thermally limited, the increase is even higher considering only thermal constraints, up to 117% of bipole capacity. The reason that we don’t seem to see a one-to-one compensation between the added DC capacity and incremental interface capacity is that there is a significant offloading of underlying lower voltage lines that eliminates the impact of certain limiting contingencies.

Coming closer into the New York City load pocket, the incremental capacity averages 76% of the bipole capacity. This is the result of an interesting locational feature: as more power is dispatched upstate, the particular power plants that are ramped down or taken offline in the load pocket causes local congestions. One could view this in another way: the bipole results in certain must-run dispatch patterns in the load pocket.

None of the interfaces studied was stability-limited, and in the SRIS process this meant that the stability limits were not pushed as much. The relative response with and without the bipoles to typical disturbances show little change. We posit that there are two influences at play here: the AC system is more stable with less power being carried on the AC lines, and it is less stable due to the reactive consumption of the bipoles. Also, we note that for the design with two bipoles of 1000 MW each, the loss of any one bipole can be withstood by the system. If the design were for one bipole of 2000 MW, this result could be significantly different.

Short circuit calculations indicate that for unbalanced fault currents the bipoles cause an increase in the total short circuit currents of some stations. Three phase short circuit currents were not affected since the HVDC delta-wye grounded converter transformers only affect unbalanced faults.

From this one example, we offer the following insights with respect to overlaying a DC line over an existing AC system:

  • On thermally limited interfaces, the incremental capacity is slightly above the DC line capacity due to the offloading of parallel paths.
  • The DC line can have negative impact on voltage-constrained interfaces due to its reactive characteristics. However, the impact is compensated for by flow re-distribution that eases system response seen from the P-V curve and could relieve certain constraints altogether.
  • Stability in general improves where sufficiently stiff connection points are chosen for the terminals. The opposite is true when the connection points have weak voltage support.
  • The DC lines themselves could become the limiting condition for stability, thermal or voltage conditions because of their high capacity. Splitting them into smaller capacity lines resolves this, but then introduces a financial burden.
  • Fault duties may rise.

One test of the above observations is to apply them to the case of the California-Oregon interface which operates in parallel with the DC tie from Oregon to Southern California. That, as they say in the textbooks, is left to the reader.  Or, for further discussion.

A more detailed cop of the SRIS for the Empire Interconnection is available to market participants and interested parties at the New York Independent System Operator.

The source for the analytical data on the HVDC line in New York is the following article: “Transmission and the Reliability of the New York State Bulk Power System, Part I: Thermal Transfer Limit Analysis,” by R. Tapia, M. Elfayoumy and R. Clayton, September, 2004, Power System Conference and Exhibition, New York, New York.