Subtransmission Planning

June, 2006

By R. Austria, Pterra Consulting

One utility's bulk power transfer, is another, neighboring utility's overloaded, under-voltaged supply path to distribution customers.  The subtransmission network (STN), as the name implies, underlies the transmission grid, as in "supports, uplifts, undermines or weighs down".  STNs are simultaneously necessary and bothersome.  Their existence allows networks to meet reliability criteria, and yet, they can also be the cause of congestions that reduce transfer capability.  Their use may lead to local cost impacts and yet improve overall system operating economics.  By their nature, STNs are akin to transmission networks in that they influence reliability, are non-controllable, in the sense that power flows naturally unless specific devices such as FACTS equipment are added, and they exist for a long time. 

Rather than be surprised by their sudden presence in our operating grids, we can apply planning principles and have some measure of control and understanding on their future influence and use.

If we name them, then we know their purpose ...

STNs are generically defined by either their voltage level or function, or both.  The typical voltage range is 34.5 to 138 kV, and a typical functional description may be "that part of the grid that interconnects the bulk transmission elements with the distribution elements."  More often, there is no specific term applied, just a part of the network that has lower voltage and has less transfer capacity than others.

They may have been designed in, as we may find in urban grids where the STN serves the specific purpose of distributing incoming bulk power to the main distribution substations.  The voltage level is selected to require less lightning and electric field protection, while still being high enough to minimize losses and voltage drop to the load.  The layout is redundant so as to provide better than n-1 reliability, usually via a ring arrangement with switching capability to transfer load to live circuits. 

Or, STNs may be legacy networks, retained after some necessary bulk system development, such as high-capacity EHV transmission overlaid on a overloaded, stressed grid.  In these cases, STNs once were the bulk transmission system and became the underlying system.  The STNs do not seem to be retired in these cases, but are kept in continuing service to provide support or as a backup system. 

The Reliability Objective  

In planning for STNs, the reliability objective as a minimum matches that of the bulk transmission system.  In fact, the STN is treated as part of the transmission network for this purpose.  Hence, single contingency (or, n-1) criteria are a staple requirement.  All other system performance measures, including transfer capability, which are based on the n-1 standard then implicitly account for STNs.  In the context of NERC class B contingencies, switching is not allowed; either the STN is normally in service or it is not.  (This leads to some unwelcome consequences as shown in the Case Studies below.)  This further implies that stability, not normally impacted by outage of low-capacity STN elements, may in fact be significantly influenced by STN configurations.

However, being closer to the load, STNs are also expected to, at least, not degrade customer supply reliability.  For distribution systems, this reliability objective is with respect to "sustained interruptions"; or interruptions to customers that last more than 5 minutes.  The primary indices are average time without service (SAIDI), average number of interruptions (SAIFI) and average duration of interruptions (CAIFI).  Fast switching, load transfers, and system re-segmentation (when the STN is reconfigured to accommodate an outage) are allowed redundancies in this type of reliability target.  [See reference 2]. 

Overall, given that one is able to calculate reliability based on component outage probabilities, then the combined reliability objective of the STN is better than n-1.  [See Reference 1].

Applying a Planning Process

To plan an STN, we start with the basic building blocks for network planning, comprising:

• A set of attributes to aim for - including cost, service reliability, environmental impact, financial requirements, etc.
• A set of physical parameters - such as specifications for new transmission (voltage level and capacity), available rights-of-way, voltage and current uprating options, component and construction costs, models of the existing interconnected system, operating procedures, and so on
• A set of future scenarios, typically based on a range of load forecasts, generation growth, distribution and bulk transmission development, regulatory structures, market structures, etc.

Once, these are defined, there are several standard processes for conducting the actual planning.  One of these is shown in Reference 3, for a least-cost integrated method applicable to deregulated environments.  The product of this process is a plan, with certain characteristics for robustness and flexibility.  The plan is a function of the companion transmission and distribution developments.

The typical analytical tools are power flow, contingency analysis, short circuit calculation, dynamic and transient stability, voltage stability, to name the basic ones.  Note that because of the switching features of STN, contingency analysis and stability analysis will need to account for these adjustments in their solutions.  Certain software may balk at the modeling of subtransmission networks because of the high impedance branches (when all branches are on the same MVA base).  An optimal power flow may be useful to conform the voltage profile during the planning process.

Case Studies

Case Study 1.  In this case study, we have a national power system whose overall load is growing at 5% annually.  It has a large urban area in which 50% of the demand is concentrated.  The urban area is served by a separate utility from the company that provides transmission service.  In the planning horizon of ten years, the transmission company plans to develop new voltage level transmission, double the previous highest voltage level (a useful rule of thumb, reflecting economics), and the urban utility needs to determine how it will develop the STN.

The feasible options, after some analysis, were:

1. Uprate some of the existing circuits to the new highest voltage level (see Techblog on Voltage Uprating, July 2005), creating an internal ring of high voltage lines that offloads the remaining STN.
2. Develop a new "hub" substation within the urban area for the higher voltage level and connect these via transmission lines on high, steel poles (to minimize the right-of-way width and EMF impact).  This is the most expensive option.
3. Sectionalize the STN to avoid parallel flows that impact their thermal loading.   This reduces the overload reliability of supply to customers because redundancy is diminished.  Some switching capacity compensates for this.  Overall, this is the cheapest option.

Option 1 was chosen for its better provision of reliability at a relatively small incremental cost to the cheapest option (option 3).

Case Study 2.  We have a legacy STN that is impacting the transfer capability of the bulk transmission system.  The impacts are thermal, voltage and stability, depending on the operating conditions.  Furthermore, the STN is owned by a separate utility from others that own the bulk transmission, although all belong to the same control area, and are operated by the same ISO.

The options are:

1. Operate the STN with certain segments normally open.
2. Implement several Special Protection Scheme (SPS) to trip a segment of the STN, switch in capacitor banks or shed generation and load.  Each addresses a specific contingency for a specific operating state for a contingency overload, voltage violation, voltage collapse or instability.
3. Add phase-shifting transformers on the STN to control flow.
4. Reduce transfer on the bulk transmission and continue to operate the STN in parallel with the bulk transmission.

The attributes of the optimal plan are a mix of both conflicting objectives such as maintain reliability and reduce costs, complex functions such as inter-utility relationships, and unbounded uncertainties such as the future operating environment.  When viewed from different aspects and stakeholder viewpoints, the best option changes.  Option 4 was selected as the path of least resistance in this particular case, although in similar situations in different locations and times, all the other options have been applied in practice.   Many of the issues that were unresolved, but "accepted" by the choice of Option 4 may have been avoided by proper planning of the STN.

Conclusions

Subtransmission networks, especially legacy ones, present a challenge to operators and planners alike.  However, understanding their nature and being able to plan for their future use helps in better integrating STNs with the rest of the grid.  The benefits can continue to be realized, while the disadvantages minimized.  A defined planning process helps in that the process itself helps understand the effects of the STN as the rest of the power system evolves over time.

References

1. Ricardo Austria, et al, "Assessing the Impact of substation-related outages on network reliability," Power System Technology, 2002. Proceedings. PowerCon 2002. International Conference.   
2. Warren, C.A.; Ammon, R.; Welch, G., "A survey of distribution reliability measurement practices in the US," Power Delivery, IEEE Transactions on
   Volume 14,  Issue 1,  Jan. 1999 Page(s):250 - 257.
3. Ricardo Austria, R. Nadira, L. Cosenza, C. Fuentes, M. Avila, and J. Ramirez, “Least-Cost Transmission Planning Considering Power Industry Restructuring,” presented at the IASTED Conference, Orlando, FL., October 1997.
    

For questions, comments and further discussion, contact me at ricaustria@pterra.us.

© 2006.  All rights reserved.