For significant use of distributed energy to arrive, it must be managed and used locally first, the variability and storage managed locally, and only then traded with others. This mode of operations is termed a microgrid. Legacy business models that assume irresponsible end nodes require direct control of energy distributed in residences. Commercial sites are treated as bulk generators subject to NERC regulations that require months of filings for each configuration change. It is time for a new regulatory and operational model.
The microgrid model lends itself recursion. Twenty home microgrids on a neighborhood street can federate themselves as a microgrid. Such microgrids should manage variability internally first, trade power first amongst themselves, perhaps incorporate additional storage as a group, and then trade with the larger distribution network. A similar logic flows up through the larger neighborhood, the district, and potentially the town.
In a similar manner, a commercial site could produce and store energy locally, manage variability locally, and trade with the larger grid only to rectify systemic shortage or surplus. Perhaps the initial microgrid is the office park. Sites and facilities with the office park then evolve themselves into microgrids, to gain additional energy surety and local control.
At some point, these microgrids that are aggregations of microgrids reach a scale comparable to the bulk generation that current NERC requirements (I’m thinking CIP 5) were written for. These include cyber-security, and configuration management and filing configuration changes way in advance. These standards are important to manage stability of the overall transmission grid. These regulations do not recognize that failure modes for these composite microgrids are quite different than for bulk generation.
Managing these composite microgrids will require changes in thinking, similar to those seen in IT for storage and for cloud computing.
A composite microgrid shares failure characteristics with a RAID array. 30 years ago, one paid a premium for disks above a certain size and above a certain data throughput. Today one pays a discount. The reason is Redundant Arrays of Inexpensive Disks (RAID), although that acronym has morphed into independent disks over the years. RAID technology multiple disk drive components into a logical unit for the purposes of data redundancy or performance improvement. By the late 1980s, it was recognized that the top performing mainframe disk drives of the time could be beaten on performance by an array of the inexpensive drives developed for personal computers.
Although RAID technology was developed for price and performance, it was soon recognized that it offered superior failure characteristics. A drive could fail without any externally-visible loss of data. A replacement drive could be added to an array with only a temporary reduction of throughput. RAID arrays properly managed nearly eliminated catastrophic loss of data.
It is an interesting side note that the first control of electricity took the insignificant charge generate by two pieces of metal separated by salt water (an electrolyte) and made it useful and predictable by stacking many such pieces of metal and paper soaked in electrolyte. This type of technology was initially called simply a pile (or voltaic pile), but was later renamed a battery by Ben Franklin, invoking an artillery battery. So it would be appropriate, albeit confusing, to say that a microgrid can consist of a battery of microgrids.
While batteries and RAID arrays offered more capacity and greater predictability, it is the reliability and failure modes I want to concentrate on here. Just as a RAID array does no fail when a single, perhaps inexpensive component fails, so a microgrid does not fail when one of its components fails, or changes in capacity. Cloud data often uses hybrid RAID, in which RAID arrays are themselves components of or RAID systems. In these, entire RAID arrays can fail without reducing throughput or availability.
An analogous increase in redundancy, availability, and resilience is the expected outcome of the aggregate recursion architecture for microgrids, or what some are calling more elegantly, fractal microgrids. Just as RAID architecture enabled data centers to incorporate inexpensive “unreliable” disk drives into mission critical systems, so reliable aggregated microgrids can be built upon small, inexpensive microgrids that are currently prices out of the heavyweight NERC-required processes.
A similar logic encompasses residential control standards. To prot the distribution grid, utilities are granted direct control of low-level devices inside homes. This results in two systems that never meet, the home-based distributed energy system and the home based energy use. Homeowners realize this out to their dismay when they have no access to their distributed generation when the grid is down.
The model of the home microgrid instead rewards the customer to install local storage. Solar installers could develop businesses around optimizing cooling when the sun is shining brightly. Such development will never happen so long as utilities commissions mandate direct control, or require a heavy process for connecting microgrids.
We need new lightweight regulatory models that embrace the coming microgrids.