These advice paths need to be separated out into those mentioned on the Main Page. They are preliminary notes, most from 2009, and are not authoritative. The most useful information on this page is the list of references at the end. Everything on this page should be considered in DRAFT and not published form.
- 1 home and small office users
- 1.1 0. baseline: 2010 home network
- 1.2 1. most important are power features
- 1.3 2. extended product line with interoperability with home grid/control
- 1.4 3. other features
- 1.5 misfeatures: wireless, no pass-through, independent power management
- 1.6 protocols; other vendors to consider
- 2 commercial users migrating to ubiquitous low-footprint BPL networking
- 3 business planning to set up unified district utilities doing power & data
- 3.1 revenue streams for unified utility (power and data) services
- 3.2 "district" utilities already exist and are getting more prevalent
- 3.3 features no other utility can offer
- 3.4 Canadian aboriginal enterprises could do this more easily than others
- 3.5 unacceptable services from incumbents
- 4 advocating to vendors to support standards and interfaces
- 5 influencing existing utilities and other service providers
- 6 influencing governments and regulators
- 6.1 clarifying "fibre" as "the solution": where it is, where it isn't
- 6.2 clarifying "smart grid"
- 6.3 "fibre to the home" is an upgrade from the basic "smart home grid"
- 6.4 three rights: G.9960 gateway meter, gigabit to TES, third party billing
- 6.5 conditions on contracts and RFPs using public funds, fixed mandatory charges
- 6.6 from neighbourhood transit exchange to transformer to meter into outlets
- 6.7 issues to debate
- 6.8 advocacy of post-DSL post-cable BPL technology, open transit exchanges
- 6.9 subsidize destruction via markets, conservation via incentives, nothing else
- 6.10 solve "last mile" delivery problems with the smart grid itself
- 6.11 push data-over-power implementation into publicly-funded buildings
- 7 references (only)
home and small office users
Home users who want reliable home AV networks without wall warts or wireless are still waiting for the perfect solution, but there are some near misses on the market already. If you're desperate you may want to just buy one and become a customer of these companies and push them to build the right thing in their next version (notably including 802.3at ports and passing through the data to female AC plug, and adding the option of UPS battery backup too).
0. baseline: 2010 home network
A well-designed network purchased in 2010 would consist of the following components:
- Desktop computers using either USB 3.0 (170 MB/s) or SATA6G (280 MB/s, striped RAID0 = 560 MB/s) as their primary boot image disk backed up by offline storage or network storage available over 1GLAN or (in offices or demanding applications) at least 2GLAN (two teamed gigabit LAN/RJ-45 interfaces with approximately 2gbps total speed to a router) e.g. ASUS P55 series.
- A router capable of wireless-N, teamed gigabit connections, external USB or eSATA disk support, FTP drive-mounting and DNS, and ideally DD-WRT open source firmware - which supports Atheros radios. The Atheros AR7400 chipset supports 802.3 GLAN, P1901 powerline communications, and wireless-N capability and should soon (2010Q3 at the latest) be integrated into routers that communicate up to 500Mbps (raw PHY rate) with every AC device, 500Mbps via wireless-N, plus an additional 1000Mbps for each ethernet port and 500-1000Mbps for each of telephone and coaxial cable. The GGL541 explicitly promises integration of all three into a 3-5000Mbps home network. eware of pre-P1901 powerline networking!
- Optionally, a RAID5 with 1000mbps performance up to 77MB/s for a single GLAN connection would serve those who have only one ethernet cable and no need for longer range wireless or powerline networking. Read reputable wireless router reviews and preer a MIMO radio (2x2 or 3x3) dual-=band combination that can do 2.4 and 5.8 at once. For more on what band to use, read a good comparison of 900 MHz vs. 2.4 vs. 5.8, unless of course you can't put up equipment on the other sie and are stuck using what the AP uses. Don't build your own antenna, good ones are cheap.
- Only devices with detachable antennas should be purchased, as these are the cheapest way to upgrade for range; The difficulty is matching the antenna to the jack on the radio or the amplifier or the router. Don't forget lightning protectors! Hyperlink/L-com can sort this out for you somewhat. Remember also that unless the antenna supports the raw frequency, and all of the radio and the amplifier (if used) and router support the band (a/b/g/n), no signal will be received by the operating system. If all you need is [a very powerful tiny b or g AP, check out ubiquiti]. For some simpler applications like vehicles with laptops, USB antennas can be used, powered via the USB cable. For antenna advice see this summary of marine antenna technologies issues. Generally you need a circular polarity or many-polarized or dual-polarized antenna for obstacles or water. Keep in mind that 400mW = 26 dBm and 1000mW = 30 dBm - so you pay much more for more power in the radio or amplifier (dBm) than you do for better gain in the antenna (dBi) though they do the same thing for you. This argues for using as big an antenna as possible.
- Power over Ethernet devices are typically used for wireless access points, and can also be useful in outage-prone areas a they run directly on DC power, typically injected at 48VDC but (for a single device) also injectable at 12VDC, 15VDC, 18VDC, 24VDC using a non-standard injector to "passive" devices. Non-PoE devices can be supported with an adapter but will obviously not gain communications capability.
- HDMI/composite media network drive capable of delivering full HD 1080p to any monitor with 5.1 or 7.1 audio and being loaded via eSATA, USB 2.0, any RJ-45 LAN, [O! Play e.g. ASUS O! Play] hooked up directly to the router or computer if either is within HDMI cable range.
- 1080p or better (1920x1200, 2560x1600) monitors, ideally pivoting into portrait mode for preparing or reviewing paper documents.
- USB or wireless-N audio headsets, with microphones to support VoIP applications
- mobile phones with "dual mode" 802.3a/b/g/n support, to use the home network for call-outs
1. most important are power features
Replacement of such capable devices will generally be justifiable only with advanced power features - every other capability (teamed gigabit LAN, 1080p playback) will not be obsoleted for a decade.
Smart power features let these devices pay for themselves - look for power bar / UPS / power prioritization and hibernation features first, these are going to differentiate winners from also-rans in this market. If you want to act on this now at home, Netgear Powerline HD Plus Ethernet Adapter Kit (HDXB111), like Comtrend, lacks 802.3at capability (requiring you to waste the pass-through outlet on a wall wart!) but has "optional manual or automatic standby functionality for efficient power consumption" to help pay for itself. Making it a power condom?'
best: 802.3at DC plugs, passthrough AC plugs
The features you really want are 802.3at ethernet plugs and passthrough female outlets that also pass on the G.9960 data, say to a laptop or desktop. Give up every other feature to get these, because they eliminate your DC wall wart. No other power feature can make up for lack of these.
2. extended product line with interoperability with home grid/control
After power features, look also for an extended product line including four-port AC and ethernet connectors like the Comtrend Comtrend PLC Ethernet Adapter PowerGrid-922 which lets you connect AV and computer bays with just one outlet.
best: convergent EoP products for UPS, SIP, AP, smart meters, home control
Look for convergent products such as a line of EoP products like SIP VoIP phones or fixed wireless APs, smart meters and other devices that the vendor guarantees will work well together and with popular home control and demand side management systems.
The very best would be a UPS in the same product line, with the highest end device in the product line being both a UPS and a multi-port pass-through AC power bar and a multi-port 802.3at router. Still waiting.
If there's a BPL power modem in the mix, you can be pretty assured the company understands interoperability. But it shouldn't have to be from the same company to work, thanks to UPA. Making the modem work with all of the standard EoP products in the home is the service provider's problem not yours.
3. other features
All vendors offer QoS guarantees for "bandwidth for multimedia payloads including voice, data, audio and video", but find impartial reviews. Netgear, as do other vendors, also promises "push-button security pairing" using 3DES (some use AES or stronger encryption) to avoid tapping into the ethernet from unauthorized points. This is a nice-to-have.
best: QoS that works, AES that works
If QoS works at all for audio/voice and video/TV, if AES is available at all, that is enough. You should be picky about power management first and product line second, and not quibble on these features.
misfeatures: wireless, no pass-through, independent power management
Avoid all products that don't provide ethernet plugs, powered or not. Avoid those with neither pass-throughs nor power features like the D-Link DHP-303 do have some features like TR-069 management for unified communication because they use the UPA standard which replaces the older HomePlug AV technology. Avoid also power management that is proprietary or requires a vendor-specific or manual control interface that can't be automated easily.
protocols; other vendors to consider
- AC: P1901 G.9960 = G.hn = EoP, DC: 802.3at
- management: TR-069 SNMP 802.15.4 IPv6 iSCSI
- services: IPTV VoIP (SIP)
- encryption: 3DES, AES
An ideal product would support literally all of the above plus others supported by the UPA.
commercial users migrating to ubiquitous low-footprint BPL networking
Commercial users need to use BPL for several compelling business reasons:
- existing AC outlets cost less than running ethernet and separate power runs
- electricity costs need to be minimized, and only EoP can do it effectively
low-footprint is a better customer value story
Using G.9960 over existing AC wiring back to the transformer, connecting to fibre either at the transformer or at the transit exchange shed/closet, has an extremely low greenhouse gas footprint. The ecological footprint of unified power and data grids is also lower than all other forms of networking, partly because the power is monitored and managed as part of the networking, partly because existing wiring is used, and because no exotic materials are required (unlike wireless antennae and some very fast chips).
Told properly, this story can be a commercial advantage in the marketplace - see the bottom-line business advantage of greener business practices. Why should customers want to pay more so you can waste power?
the office you want to work in
Another important advantage is to attract the best network-concious staff. In the modern vision, storage/processing hardware is gone. Invisible. Storage and processing come from "the cloud" in which there are standard interfaces to marshall more processing, more storage, on some hosted platform. If you ever care where the host is, you've made a mistake.
Users must be able to walk into a room, point a mobile device at a screen (or at the very worst plug in a USB key) and watch it all be customized to your preset needs and permissions. "The computer" as a set of devices comes into existence only for that moment for that purpose, and evaporates into separate devices on common (ethernet or ethernet-like) backplane waiting for someone else to use them in some different configuration next time. It's an incremental change from today's USB devices swapped in and out often. In the post-USB future that will be paid for with electric power savings and shiftings, and is already being deployed in poorer places, each AC outlet has a networking device on it, ideally one capable of providing DC power on a powered ethernet port. Like USB, AC plugs and DC devices on ethernet cables get power and data from *one* cable. Wireless use for immobile devices will fade in favor of this more secure faster model that leaves wireless spectrum for actual mobile user needs.
the office power system 2010-2030
How you connect to the "backplane" depends on what kind of power the device needs:
- AC devices connect via a wall outlet using the G.9960 protocol over UPA compliant plugs
- DC devices connect via a powered-ethernet 802.3at RJ45 capable of supplying 60 watts
- battery or solar-cell-only devices connect via 802.4.15 (a very low overhead IPv6) or WiFi
All these devices will have IPv6 addresses and eventually iSCSI interfaces like hard drives. At that point, the entire home area network on your side of the transformer on the pole is "a computer". Most office users will simply have an efficient PCI-E rackmount slid into a wall somewhere - and home users will almost certainly just rely on laptops/notebooks and some big screens. If you have a "blade server", it may have a faster backplane like 100 Gigabit ethernet (the current PCI-E bus can handle 40 gbps already so the next upgrade is to this faster backplane speed developed for use in telecom switches).
No rival architecture will be emerging for 15-20 years. It takes that long to come up with a standard for AC or DC or backplane card devices. We're going to be as stuck with the above as we are now stuck with PCs/phones. It's a "plateau" that can't improve without abandoning so many existing AC and DC standards that it won't be happening probably within our lifetimes.
total cost of operations
Drawn primarily from a CEG analysis of these issues for City of Toronto in 2004-5 especially an exhaustive analysis of TCO factors, that highlight the importance of buying long lived equipment at a tech plateau.
Baseline best practices "include longer lifecycles on equipment... use of parts that a single 'smart' driver for the entire product line...NOC-based "installation" and "re-imaging" and "cascading" procedures, removable boot drives - easily re-imaged (without a deep diagnosis) and removable data drives which can be easily moved to other machines and isolated..." and etc., in addition to the power management considerations dealt with above.
business planning to set up unified district utilities doing power & data
revenue streams for unified utility (power and data) services
The only way to pay for a smart grid, in those more sane jurisdictions that do not permit the powerco to put fixed charges on power bills to pay for infrastructure, is (in the order of the public's willingness to pay for them)
- negawatts (someone pays you for a watt you saved them to use elsewhere, the customer gives up a 5 cent kilowatt to save someone a more expensive one and can get a portion of that savings back)
- climate offsets (received by the powerco for any savings, esp. peak)
- voice service running on the same communications as the smarter home/grid
- TV service running on the same communications as the smarter home/grid
- data service running on the same communications as the smarter home/grid
- jacked-up per-kilowatt rates (always good, encourages conservation)
The mobile device will be an open platform like Google Android, and there will be no way to prevent it from using VoIP or data services from any third party that can run a wireless access point (AP) out of any back window. A lot of new phones can accomodate two or three SIP connections over WiFi in addition to their GSM network.
It'll be pretty normal by 2012, standard by 2015, to buy only open phones and program them to find the best deal for any given call from one of two or three networks. The monopoly will have broken down to the call by call level. There will be literally no advantage to being the incumbent carrier.
In the meantime a lot of money will be made offering long distance cordless (using equipment like Senao) or WiFi phone services to those who'd otherwise use cell phones over GSM, and pay for every minute. More revenue.
"district" utilities already exist and are getting more prevalent
There are many existing community co-operatives, apartment complex or industrial parks, and other entities like native reserves, that could or should take control of their utilities. Putting in their own electric storage & generating devices is a popular idea but it's an economic loser if they're still relying on the monopoly powerco and communications companies to carry the data and optimize their usage. It just can't be done properly above a district / neighbourhood / village scale. On the largest scales, cities like Miami are relying on Silver Spring to unify the power and data with many underlying service and equipment providers.
features no other utility can offer
Typical services deployed include:
- "“in-home energy displays or “eco-panels” to help manage electrical loads and lower power use during peak periods,”
- “smart appliances that can communicate with smart meters to reschedule high-energy functions or switch to a lower-consumption mode during peak demand periods”
- “programmable and smart-meter-controllable thermostats”
- “demand management and demand response software that will manage consumer appliances, lighting and other devices using smart meters.”"
If the utility is contracting to provide not "power" as such but heat, lighting, DC power to the device, refrigeration/freezing, etc., then it would be up to the utility to decide how to deliver those services. Write the contract correctly and let the utility own and update all the power-using and power-wasting devices in the home. So a lot of crappy electric area heaters disappear and a few basements get insulated and that's paid back over some years on the bill, while some always-on devices get replaced by smarter ones, with the utility itself managing the bureaucracy of qualifying for the grants and rebates, and the utility itself keeping some portion of the savings...
Canadian aboriginal enterprises could do this more easily than others
When considering a unified district utility for a reserve, consider that:
- the band owns the land and has all the rights of access required to fix or upgrade things, including often the housing itself
- aboriginal enterprises get favourable tax and credit treatment especially in economic stimulus programs
- there's lots of aboriginal youth who understand this tech extremely well and can learn what they need to know in a very short time, qualifying for job training money and eventually wiring up all the nearby communities
- aboriginal culture money can perhaps provide seed where local TV and/or radio stations are being set up to run on the unified district utility
- aboriginal housing is in crisis and any excuse to get in and fix stuff that is obviously wrong could improve things - that is while insulating one might find mold problems, while wiring one can find and seal to exclude bugs, etc.
Also there's the question of autonomy and developing basic economic capacity:
- bands using a utility they have a stake in, that runs their own encrypted communications including wireless, building capacity for business services
- the utility could offer reserve-to-reserve secure links to enable high level conversations fit for nations to discuss with nations
- best practices can spread relatively rapidly to other reserves nearby and there's a natural market affinity and service market in advice and guidance
unacceptable services from incumbents
Ultimately, as with governments, one ousts a utility for simply not performing:
Some reasons why customers might prefer to oust an incumbent powerco or ISP:
- fixed charges to pay for infrastructure that doesn't benefit the customer, like non-IPv6-non-G.9960 not-smart meters
- attempts to lock out players like Silver Spring, IBM, GE, Google who all have services to unify information from various devices and show them to the consumer, that are independent of the hardware and bandwidth between
- failure to cooperate with established home control players (like Zigbee)
- giving up data indiscriminately to major corporations like Google without explicit customer permission
- meter data latency which could be up to 24 hours in the worst case
advocating to vendors to support standards and interfaces
Anyone ready to buy lots of equipment, or with needs similar to those of many others who would buy equipment, can advocate meeting standards and interfaces. Often vendors are quite open to receiving this kind of input.
what to ask vendors and integrators for
Business users should be contacting Netgear, Comtrend, Corinex and others with a broad product line, and APC and other UPS makers. Suggesting that they should perhaps merge and join with a midspan router maker may help the desirable device emerge faster: the UPS, 802.3at midspan and G.9960-compliant home AV network in *one* device that still looks like this. Without this one device, power management is split among the power condom on the wall outlet, the small device that uses only DC, the 802.3at router powering your VoIP and wireless AP and so on, and the UPS.
While unified power management software (from Google, GE, Zigbee, Silver Spring, Comtrend, EnerNOC, etc.) may relieve this someday using TR-069 or some other standard, for now the safest bet is relying on one vendor to smooth out the kinks. Interoperability tests at major events (like Interop) may help discipline the market. So you should ask which they participate in.
it's a green decision: tell the vendor that
Finally the vendor and integrator should be asked to explain in one single spreadsheet how much power and waste will be saved by using their solution, and make clear that this is how you expect to pay for any networking upgrades now and in the future - that the purchasing decision is a "green" one and you are looking for maximum equipment lifetime and low total cost of operations. RFPs must especially demand this as part of the conditions of responding.
More on this issue at ask vendor to support standard or interface.
See green networking purchase for how to write state of the art RFPs.
influencing existing utilities and other service providers
If you aren't going to set up your own unified district utility, and very few will, then nudge the incumbent to move faster. One can hint at changes to regulatory regimes already undertaken in other jurisdictions, questions can be raised "about the way some utilities inhibit efficient power demand management by providing no way for it to reach into every AC outlet" and the "innovative communities" already doing this can be listed (don't forget to name Silver Spring, Cisco, Google, GE, IBM).
Some will be reluctant, forcing you to apply pressure. You need to at least make an honest effort to pay for your smart meter and power grid maintenance by deploying bandwidth. And making the in-home over-AC network plug 'n' play compatible with the power modem that is distributed for BPL users.
getting bits into the home over power lines (BPL over LV/MV)
If you're new to the idea of ethernet-over-power (EoP), home grids, and BPL, then look at this first. Broadband connection over power lines solves a lot of problems that fixed wireless - like Motorola Canopy - and DSL and cable cannot.
There are many reasons to prefer using power lines to actually get the bits right into the home, notably integration with the smart grid and smart meter systems.
With end to end BPL done right, especially by the powerco itself, electricity demand management pays for broadband deployment. See recent news in powerline and IPTV, the primary application.
leaders and laggards
Over 2010-12 the competitive playing field will be established with different Quicken-like services (GE explicitly compares it to that in TV interviews) giving you a view of whatever devices can participate at whatever level of control detail, setting power management policies. You won't be able to buy a UPS or G.9960 router or WiFi access point or powered ethernet switch that doesn't have at least the ability to participate in these efficiency networks. Most such devices will be *sold* based upon the *power savings* just as light bulbs, convection microwaves, power bars and UPS are already. The smart utilities [ http://earth2tech.com/2009/05/20/utility-perspective-why-partner-with-google-powermeter/ will say]
- “To pretend that we’re going to completely own that relationship with the customer and not work with companies like Google is naive. It’s the customer’s data, we should be seen as a facilitator.”"
In more primitive jurisdictions you will be fighting with nuclear contract overruns, fixed charges on bills, and still stuck with the duopoly of cableco/telco. abusing their power to deploy the dumbest possible meter system and impose the most overtly abusive and monopolistic terms on the access to the customer's bill and service cable/meter. The people who'd go along with that are the 60-somethings who can't learn the new way... and they'll be severely crippling the economy of anyone forced to rely on them for power, who will probably have to route-around using Comverge or EnerNOC or equivalent who will get access to customer bills and raw data as long distance competitors got in the 1980s.
which type is yours? leader or laggard?
The most important thing to figure out is whether your powerco is a leader or laggard. If it's a laggard you have to figure out why and deal with it:
- some powercos still think like broadcasters with a control of supply, and refuse the right to access a pole and access meter data, as long as they can; these have to be pressured by business users of the UPA standards and operators of many buildings and municipalities: You literally can't roll out electric cars or get rid of coal burning to meet peak electric demand without a much smarter grid that reaches right into the home.
- telcos and cablecos may have managed to make the powerco dependent on them; in the long term telcos and cableos will be forced out of last mile connectivity and certainly also as providers of PSTN connectivity and TV/radio access. They've built their business model on arriving first and squeezing or arriving second and charging slightly less for slightly better service, and gouging governments to pay over and over again for the same old wires; powercos may have to be encouraged to shed dependence on these first
- within the telco you may have to spread comfort with the idea of giving up the last mile and living in service closets like the Vancouver Transit Exchange already has, trying to keep customers with better video-on-demand service or 100,000 satellite radio channels from all over the world delivered over broadband or amazingly good services combining mobile/cell and home data or robust successors to the "I've fallen and I can't get up" type services. These are good businesses, actually better than land lines. Telcos will still do backhaul and peering at major IXPs, but increasingly over dark fibre that the public owns and controls and tariffs to them at the same rates as other providers (or that they own but are forced to tariff).
- within the cableco the paths vary: they already use the ethernet-like tech the powercos will be using, and already understand how to go up on poles without being electrocuted, so may become contractors maintaining that last mile around the service closets, and they're in a better position to run a fibre backhaul based on standards than telcos are (which invested in DSL). But they don't understand reliability as well as telcos or powercos so I expect they will ultimately fall back to running fixed wireless links and packaging specialty TV and video-on-demand channels, and wiring up the institutional and commercial clients that already have LANs and wish to outsource dealing with spam and malware and botnets and general operations.
Most dominant players in dying industries do not survive because they can't see the obvious. By 2015 the survivors who held out longest will actually be in the *worst* shape, because they can only survive by abusing their power to deploy the dumbest possible meter system, keep others off the poles, and impose the most overtly abusive and monopolistic terms on the access to the customer's bills and service cable/meter... severely crippling the economy of anyone forced to rely on them.
the powerco's inherent edge: paying for provisioning with efficiency
Telcos and cablecos are afraid of aggressive enlightened powercos for good reason: A powerco can amortize its provisioning over a shorter period (or more intelligently, amortize *more* provisioning over the *same* period) because it is able to assume very high cents per bit payback in its energy management (negawatt, watt, peak levelling) and safety monitoring (you're already watching power, why not watch gas and water and fire sensors). Since these require even higher reliability than the PSTN network, adding voice will cost nearly nothing. Then additional lower-security less-hardened networks can be used to carry TV, and a wide open least-hardened open Internet connection can run on leftover bandwidth.
Telcos and cablecos can do *none* of this. A powerco or third party power management company (which the powerco will be forced to accomodate under a tariff) will only contract with a telco or cableco where it's not worth it for the power player to provision themselves. However, remote areas where there is no wired broadband at all, could get provisioned *first* due to the $100/month I'll happily pay for even relatively slow (but wired) low-latency broadband. So BPL pilots grow by eating satellite, then fixed wireless, then DSL and ultimately even the areas served by fast cableco networks now.
When the difficult problems of serving the boonies are solved, by trials like IBM's on Washington Island, Wisconsin, the same tech rolls out to the nearer rural areas still relying on satellite and fixed wireless, and then ultimately creeps ever closer to the cities...
influencing governments and regulators
It won't be particularly easy to get governments to understand the new way an open combined power and data grid must be regulated and forced into level playing fields by their decisions. Most politicians know that "we are not smart, we cannot make it go" and will not risk technical decision making, even if failing to make a decision dooms the entire jurisdiction to permanent backwardness. Letters to governments and regulators have to keep this in mind.
When they do, communications companies will have to compete directly to manage your power. Why? Well, when the cable companies started doing phone ... the phone companies did tv ... and they all now do both plus Internet. So when the powerco starts selling data connectivity, because it has to, the others must be allowed into the power management market.
clarifying "fibre" as "the solution": where it is, where it isn't
Fibre is expensive and can't carry power. Wireless is flaky and insecure. These are not solutions when we want to replace universal secure systems like PSTN or support genuinely optimized electrical configurations. Fibre to neighbourhood transit exchange sheds and BPL into the home (whether fibre is used between poles or not) with a smart meter as the G.9960 gateway, with a gigabit guarantee to local transit exchange as the key regulatory goal. If delivered over BPL, that competitor can pay for itself with both negawatts and moved-off-peak watts. Fibre has no such capacity unless it's bridged into every AC outlet - again by use of G.9960 in homes.
clarifying "smart grid"
The home gigabit network and the smart grid are the same solution - G.9960 over existing AC power wiring in the home (EoP for 'ethernet over power').
Then every AC device is on a gigabit network to a trusted transit exchange shed/wiring closet/IXP where it can participate in a very long list of essential emergency and efficiency services. If the homeowner wants to pay for fibre from that TES to their home, great, but they'll also have to pay for any additional backhaul that they use from that shed or closet to the open Internet. Anything else just subsidizes bittorrent hosts and punishes those who only really need a gigabit for reliable voice, power and basic data services like email and web browsing.
"fibre to the home" is an upgrade from the basic "smart home grid"
Using G.9960 smart meters, powercos have to participate as just one of many providers of certified, authenticated, secure, encrypted AC powerline based data services. Using PoE devices plugged into AC outlets cuts power use more by eliminating wall warts. Users could and should have the right to run fibre for less secure, unregulated LANs of their own that terminate in a fibre optic connection you have paid for back to the shed/closet - a luxury or a business expense.
As a public policy term, "fibre to the home" must go away. It's not the essential feature of these solutions. The essential features are dark fibre tariffs between the poles, smart meters that are also reliable gateways for other secure authenticated services to a gigabit network to every AC outlet, and open access at neighbourhood transit exchange sheds for trustworthy hardware & open billing that maximizes customer efficiency options. If that's what's understood, then the term "green broadband" works as well as any.
three rights: G.9960 gateway meter, gigabit to TES, third party billing
To achieve the optimal energy and communications access, government must guarantee only three rights to the utility customer / homeowner:
1. A secure G.9960 gateway that is also a power meter on existing AC wiring will be installed in your home and paid for by the power savings that it enables, via a powerco-run or third-party-run (EnerNOC, Comverge) peak curtailment program, or for much higher rates for on-peak power since you can easily avoid using it.
The powerco will have to be literally forced to do this. They like their monopoly and most do not yet see DSM or peak levelling as a profit centre, as others do.
2. Regardless of how the signal gets to the neighbourhood TES, building owners have a right to access any qualified service provider willing to put equipment that meets certain high standards, on that LAN-speed network established from home to TES. While regulations should be agnostic on how the gigabit guarantee to transit exchange is achieved, doing it with BPL has the additional benefit of allowing shared batteries to keep VoIP running in outages, supporting electricity co-ops with shared windmill or solar cells, power rationing to protect vulnerable persons, and other ways that pay for the network itself. So this should be the baseline access for all TESs but if someone wants to *also* pay for a fibre connection from TES to home, let 'em. It's their right.
The telco and cableco must be literally forced to participate at the TESs and to migrate from their own legacy architectures. Their ownership rights to any legacy cable from this TES into each building must also have a sunset schedule - after a certain point they'll be owned by the community or collectively by each building owner. This is exactly how you ended up owning all the wiring inside your house or apartment building, which use to belong to these monopolies. After a certain payback period, it belongs to the owner(s) of the building(s) and becomes a "tail" in exactly the sense Bill refers to in his "homes with tails".
3. Any third party that wants to pay for a conservation upgrade, including say new fridges, and then monitor them and invite them to participate in peak curtailment (e.g. shut down for an hour when peak power is costing the utility a fortune or generating a lot of pollution), has guaranteed access to power billing to ensure they are repaid. The powerco itself can't be allowed to restrict access to the bill or list of customers.
Again powercos may fight but some like Toronto Hydro or the Berkeley CA power utility already do this kind of thing. Even Nova Scotia Power administers conservation programs for the province (which is admittedly a conflict with selling watts, and they are trying to get unjustifable fixed charges from customers to pay for it...!). But what's really needed is an absolutely flat playing field in conserving so that anyone who sees an opportunity can offer to replace wasteful old devices with new efficient ones, and receive reliable streams of data and paybacks good enough that they can go to investors and get financing to replace all the old junk in our homes with fancy new junk that sucks way less power and turns off like a good grid citizen.
conditions on contracts and RFPs using public funds, fixed mandatory charges
Public funds paying for contracts or RFPs for any services (to schools or hospitals, for instance) must be stated in terms of the above conditions, and require other important guarantees like latency, power reliability, what types of services have an absolute right to low latency or lack of throttling, what types of services do not and when they can be curtailed, and how exactly disputes with powercos, telcos and cablecos are handled and what role government plays. In Australia the government had to commit to own a national broadband network mostly because these were very complex.
- the right to access fibre on the poles,
- the right to form local electricity co-operatives,
- full power grid ownership
This last guarantee, eventual public ownership or free access to wires they pay for, is already offered to private building owners who own the wires in most jurisdictions after a payback period no matter who installed them. By contrast, governments often subsidize the extension of telco and cable infrastructure. Australia, to escape this trap, has committed to public dark fibre and reliable stable tariffs for access to it via their "National Broadband Network". US policymakers are moving in that direction too. The reason is that public agencies and utilities are going to pay several times over for the same fibre if they do not: for their own telecommunications including VPNs for schools and hospitals, to guarantee universal wired broadband to rural areas, for real-time monitoring of electric power grids and devices participating in home grids, smart grids & DSM.
Obviously, also, governments need even more robust and extensive versions of the same total cost of operations case and green purchasing case that any other business does. They should be setting standards for private corporations in this respect.
See green networking purchase for how to write state of the art RFPs.
from neighbourhood transit exchange to transformer to meter into outlets
The more we reinforce the duopoly, the less likely we are to ever get to the open TES where "homes with tails" terminate their BPL (which must be the standard, see below) or fibre (if you pay for it, and you'll still need BPL). Telcos in the US were even able to restrict access to their equipment closets due to foolish deregulation - a recipe for disaster.
What such TESs look like in practice is a single termination point with both BPL and fibre connections, and also the ability to link existing telephone wire or coax from any home. Exactly like telecommunications closets in well-provisioned office buildings. Responsible service providers can physically access the closet. Ideally any of them could put host hardware, such as video-on-demand storage or switches capable of interaction with wireless or wired backbones, or battery backup. A few standards like PoE and 48VDC power (long the standard for third party hardware in telecom closets anyway, PoE injects DC power at 48V) apply. Community co-operatives could form and literally take control of the closet just as they could in a co-operative housing development. Emergency and safety services like fire, security and control of power storage/distributing/generating devices, even medical monitoring, could run on this more secure and managed neighbourhood LAN, and implement local policy like power rationing in outages to ensure the most vulnerable persons are safe and essential services continue.
issues to debate
Issues that will come up in the process include
- so-called "smart grid" power management systems on MV and LV wires on the powerco side of the transformer such as those deployed by Aliant XWave on many Canadian power grids like NL and NB or incorporated in Ontario's policies
- demand side management systems, so-called home grid systems, based on G.hn BPL infrastructure, TR-069 and P1901 metering, delivering a gigabit to AC outlets - these pay for the provisioning, e.g. Australian Minister for Broadband speaks about broadband and climate change, 2009-04-28 who argues that electricity demand management pays for broadband deployment
- so-called "cloud computing" processing, such as Amazon's, IBM's, GNU's and so on, that allow secure VPN connections between processors and outlets to provide storage and processing power anywhere on Earth.
- district transit exchanges (Wikipedia:IXPs with commercial exchange functions) that permit users to easily choose between very many providers of services, with guaranteed bandwidth, latency and reliability back to the LAN they control
advocacy of post-DSL post-cable BPL technology, open transit exchanges
Examples of the kinds of advocacy that this grid wiki enables:
- letter to NDP-NS regarding smart grids, 2009-05-24
- letter to NS NDP government re rural broadband, 2009-06 and opposition
- letter to NS DoE regarding smart grids, 2008-01-11, in response to call for public input
- letter to officials re wiring isolated but dense communities with broadband
- MODL public information sent to residents re universal broadband, 2009-05
- letter to NB Premier Shawn Graham re BPL pilot project, 2009-05
subsidize destruction via markets, conservation via incentives, nothing else
Some arguments about subsidies:
- policies to subsidize supply of electricity instead of rewarding actual destruction and conservation *more* than continuing to generate and move watts are counter-productive and should be replaced by methane destruction, nitrous oxide reduction, smart appliances, power condoms and peak levelling
- focus on "renewable" generation instead of conservation results in rules that actually inhibit methane destruction projects and other greenhouse gas removal by stealing the climate credit for these to pay to increase supply of electricity
- putting greenhouse gas emission destruction/prevention/reduction first and conservation second, ahead of any levelling, and those all ahead of any new generation capacity, is the only order of priorities that reflects physics
solve "last mile" delivery problems with the smart grid itself
- an MV/LV delivery platform that delivers broadband between the poles (how ever it gets to those poles) can perform better than Motorola Canopy, be cheaper than DSL and deliver SCADA information as part of network maintenance - this tells exactly where faults are, etc., and pays for itself partly with averted maintenance and averted line loss
- a power company deploying broadband to consumers can easily pay for smart meters, to advance credit to upgrade to smart appliances, with bandwidth charges. So the broadband users pay to smarten everything else up!
- by no means should powercos be allowed to *charge* for "smart meters" that are not even IP-integrated and have no capability to act as a gateway (like the "power modems" in use on the island) - they should be forced by tariff to install them and recover the costs from electricity savings from smarter device management and direct bandwidth charges
- powercos are already heavily regulated and used to universal service - there's no reason at all to have a land line if the powerco implements a properly run VoIP system - most people would prefer that plus a cell to a PSTN land line
push data-over-power implementation into publicly-funded buildings
The most likely way policymakers are going to "get" the connection between power management and universal connectivity will be to use powerline networking at work and home and see for themselves how much simpler it makes electricity conservation and peak curtailment and emergency shutdown.
"Optional manual or automatic standby functionality for efficient power consumption" is only the start. Eventually these devices will be able to:
- act like a much smarter power bar or UPS in concert with other devices they communicate with, which should eventually be everything in your house
- replace the "wall wart" with an "802.3at compliant" powered ethernet port, providing 1.5 to 60 volts of power, up to 80 watts, with a gig of data on the same one cable (a cat5/6/6e ethernet cable) to devices like wireless access points or VoIP phones
- actually replace the UPS and ethernet router you are using now by fusing together the DC power, UPS, router and phone jack functions in one device
By 2011 that's what we'll all have in our homes and we'll wonder why we ever tolerated a separate jacking system for phones, cable TV, AC power, and why we ever tolerated DC "wall warts" at all. There may be a USB port on the device so we can plug in a USB key to identify who the end user is. Then the TV, computer, network settings, lighting, can all adjust for that person. Great for kids, seniors and those incapable of choosing settings.
At that point the idea of extending powerline networking into other public places, the power grid, schools, etc., will not be novel, but obviously right. So encouraging use of even primitive early UPA devices from Netgear, Comtrend or Corinex can make it simpler to achieve the other outcomes.
- World Economic Forum report on accelerating smart grid investment, underemphasizes emergency and ordinary communication piggybacking, but does mention "secure and ubiquitous communications" as benefit in every case
- P2P versus global warming: Arnaud, Hubley comment
- IPTV coveraged by DS2
- Intel standards case
- Comtrend specsheet for PowerGrid922, add 802.3at ports and a UPS and that's the thing
- Google ecommercetimes story regarding Power Meter and Google's smart grid play
- Cisco smart grid play
- Miami rolling out a million meters with GE, Cisco and Silver Spring
- New Brunswick pilot project in peak levelling, involving about 750 buildings
- grid hacks! easier over wireless than any other way - why wireless will not be the way we are networking power meters in the year 2014
- Big opportunities for ESCOs in demand response market, Smart Grid News January 20, 2010, lists 65,000MW DR potential
- OpenADR report and collaborative - supporting Smart Grid Interoperability Standards Framework which includes OpenADR 
- OATT (open access transmission tariff)
- OpenADR in Southern California, Honeywell estimates 44,000 MW now under DR, 188,000 to go, a much higher estimate of the potential than Smart Grid News reported (65,000)
- US NIST smart grid front page lists US funding etc - Obama put $3.9 billion into this, Canada has put nowhere near 10% of that in (guess why http://www.guardian.co.uk/commentisfree/2009/aug/24/climate-camp-canada-oil-tar-sands ).
- The original US policy announcement from Depts. of Commerce and of Energy:
- The Smart Grid Interoperability Standards Framework - nothing will connect to New England or anywhere in the US unless it meets the relevant standards; The OpenADR collaborative recieved many comments on it