Smart Grid – System Communication Architecture for Residential Customers

The greatest opportunity to realize benefit, yet the highest risk in terms of recovering costs, lie in the deployment of the Smart Grid to residential customers. Much of the Smart Grid-related functionality has already been provided to the commercial and industrial sectors, as the economics of higher revenues and lower deployment costs (relative to delivery capacity) create a strong business case with little, if any, risk. Therefore, in considering the dynamics of communication system architecture and its role in analyzing the costs in Smart Grid deployment, this discussion will narrow its focus to the impacts of these types of decisions on residential customers.

Smart Grid Communication Costs

Figure 1 (below) illustrates the relative distribution of capital (CAPEX) and operating (OPEX) costs for the major elements of the typical large scale residential smart grid network.

The core network is a critical element in the solution, as most of the system intelligence, databases, user interfaces, operational control, data management and security mechanisms reside here. However, it is generally the least costly element of the network and, if properly planned, is most adaptable to future changes (from a timing, cost or visibility perspective). Thus this portion of the system architecture requires the least scrutiny in terms of assessing deployment costs or risks.

On the other hand the costs of the access network and residential devices dominate the overall cost profiles (CAPEX and OPEX) of the system and the consequences of selecting an inappropriate technology or product is staggering. Consider an electric utility with 300,000 smart grid customers. The network would consist of 300,000 residential devices, perhaps 200 to 1,000 access nodes and one or two core networks.  One can reasonably replace components in the core if the correct design or products are not selected. However, replacement of equipment in the access network or customer premises it unfathomable.

Access Network and Residential Devices

The access network and residential devices are typically viewed as a mated pair (see Figure 2). They must “speak” a common language (i.e., protocol) as very few access networks and devices are multilingual. For example:

• WiMAX devices and access networks speak to one another over the 802.16e wireless protocol using equipment that meets the WiMAX specification.
• A proprietary 900 MHz access network will only communicate with a 900 MHz device manufactured by the same manufacturer as the access network.
• DSL modems speak to their access network over a twisted copper wire pair per the ITUG.992 family of open standards.

However, the above three technologies will not inter-operate. The three distinct protocols preclude these devices or access networks from communicating with one another; and similarly even within a network, the device and access network must be coupled and considered in tandem.

The challenge confronting each electric utility is to determine which access network/device combination, i.e., technology, is best suited to provide a cost-effective and reliable communications network for smart grid.  Studies completed to date indicate that there is no single technology that is optimal under all circumstances. The suitability of each technology is largely dependent on the electric utility’s system requirements and local conditions in its service territory.  As a result, every electric utility must conduct a study in order to select their unique optimal technology.

Network Architecture

Wireless Networks Look Most Attractive

Wired and wireless networks are both viable alternatives, but with vastly different deployment costs. Wireless network build costs can range between $120 and $250 per customer depending on density of customers covered; wired network costs typically vary between $500 and $2,000 per customer. This would suggest that for electric utilities that plan to build their own networks, wireless networks offer the only viable alternative (noting however, that these average costs for wireless networks increase dramatically in areas with hills, mountains and a high concentration of tall buildings).

Existing Wired and Wireless Infrastructure May be Available

Electric utilities should, however, consider that public telephony carriers have already invested in wired and wireless systems and will likely offer access to these networks at compelling rates. While these networks may not provide ubiquitous coverage in the electric utility’s entire service area, it may be to the electric utility’s advantage to lease network capacity instead of building it. The market has not matured to the point of establishing a well defined market price for leasing network capacity for smart gird; thus, the price can be expected to vary widely between carriers. Furthermore, carriers may not have networks in all of the electric utility’s territory, but may seek to leverage the opportunity to expand their coverage. In these cases the carrier may seek a long term lease agreement in return for the commitment to expand their network. Some carriers have even suggested that the electric utility might provide capital to build networks where they do not yet exist and lease the network back to the carrier.  Again, the alternatives are numerous and will largely be dictated by local considerations.

The Choice Between Wired and Wireless Communication Networks

The choice of Smart Grid communications technology is best made in the context of local conditions, electric utility requirements, and potential partnerships with carriers. In many areas wireless technologies will be optimal.

Advantages of Wireless Networks

Some of the advantages of wireless networks include:

• Several wireless networks are typically available in most markets and they can provide highly reliable service.
• They can be quickly expanded at modest cost and the wireless industry is generally open to wholesale leasing of network capacity to utilities.
• Most of the networks are built to high availability standards that are similar to the operating standards of electric utilities.

Drawbacks of Wireless Networks

Some of the limitations of wireless networks include:

• No wireless network covers 100 percent of the territory.
• Many are built to 90 or 95 percent coverage reliability, meaning areas within the service territory will have a weak signal or no signal at all.
• Companion technologies might be needed to cover all residences in an area.
• Problematic for meters that are below grade or in confined spaces.

Regardless of whether a wireless network is built or leased, a thorough understanding of meter locations and local topology is critical to analyzing the practicality and cost structure of using wireless as a smart grid access network. A radio frequency coverage analysis can be useful in providing accurate cost estimates for building a wireless network.

Key Characteristics of Wired Networks

Wired networks have the potential to overcome some of the limitations of wireless networks, but also have limitations. These limitations include:

• Cable and DSL systems are only deployed in areas that have high population densities where the enormous cost of deploying fiber, wires and coax can be justified.
• Typically limited to urban areas, the coverage areas for these systems are typically much smaller than wireless.

Broadband over Power Line

Broadband over power line (BPL) presents yet another viable technology. The most promising architecture is a hybrid of wireless and BPL.

There is No Single Communication Solution

In summary, a communication technology that may be very promising in one region may not be suitable for another.  Consequently, the optimal access network and residential device technology must be based on a thorough analysis of these local factors, and it is highly likely that an electric utility will have a mix of access technologies.