Acciones de desarrollo formativo

!

Fiber Network

• There are 152 miles of streets in the city and all of these streets would require fiber

construction. The city today has about 30% of existing utilities buried and 70% above ground. However, Austin Utilities is in the process of burying more utilities over the next decade and it is expected that by then the percentage of aerial and buried utilities in town will be about 50/50. This creates a dilemma for designing a fiber network now. It would be costly to place the utilities on poles now and then move them underground within a short period of time.

!

We elected in our assumptions to build half of the network in the city underground and half aerial. We would basically go ahead and put the fiber underground on those streets where that is already planned. That ends up costing more now since we have to trench in those streets, but it is far less costly than building an aerial network and then moving to underground later.

!

• We got estimates of road miles for the rural addresses in the Austin Public School District

from Jones, Harbaugh and Smith, a civil engineering company that works for both counties. They said that there are 89 rural street miles in Freeborn County in the Austin Public School District and 191.5 miles in Mower County. In our study we estimated that fiber would be built to 92% of these miles. There generally are a few stretches of road

that don’t need fiber, such as a stretch of road past the last house on a given road, or of there are multiple possible routes to a given home and we choose just one of them. We have assumed that all rural fiber construction will be buried and that there also will be buried drops. There might be some savings to be achieved by putting some of the rural network on poles, and so our assumptions are probably conservative.

!

• Austin Utilities owns all of the poles in the city and we have assumed that the fiber

business would lease space on these poles at market rates. In the process of allowing the fiber business onto the poles the business is subject to what is called make-ready. There are national standards for how different utilities share pole space, and to the extent that any of the existing poles are too full then the new fiber business would be required to pay for a new pole to replace the existing one. The fiber business also would have to pay any cost to move wires of other utilities if needed to make room for new fiber.

!

• There are issues to consider with having any of the network on poles. Any utility on a

pole is subject to weather problems and subject to damage from storms, broken poles or fallen trees. But a fiber network on existing poles would have the identical issues of the existing utilities on those same poles except that fiber is tougher and harder to break than other kinds of cables.

!

• Our study assumes the use of pre-connectorized fiber drops. This a relatively new

technique that consists of installing connectors into the feeder fiber and then using drops that are pre-manufactured to plug into the fiber. These drops speed up installation and repairs since no splicing is needed to install a drop. Splicing requires an experienced fiber technician and bulky equipment while any tech can plug in a drop. Our cost analysis shows that the pre-connectorized drops are not only more efficient but would cost less for Austin. It’s also easier to replace these drops in the future should they be cut.

!

• In the rural areas we will also use pre-connectorized drops where it makes sense, where a

home or business is relatively close to the road. But for buildings further off the road we would instead construct the drop using traditional splicing.

!

• In looking around the city we estimate the average drop length to be around 120 feet. The

average length of a utility connection for electric is 70 feet. But electric service generally connects to the closest point to the street while the telephone and cable line often connect to some other point on the home. This is a figure that we normally see in larger towns and reflects the relative density of Austin compared to many other similar towns. We have estimated that business drops in the city are 220 feet, reflecting that some businesses are offset from the road due to parking lots. We estimate the rural drops to average as long as 300 feet. We have assumed that all of the drops will be buried and that they will be plowed rather than bored.

• We estimated the number of residential households in the city to be 10,600 housing units. There were several sources of data showing the number of residential housing units in the city. For instance, the Census shows 10,890 residential housing units estimated for 2012. There are also estimates created locally by Austin Utiltiies and there also was a housing study done by the City of Austin that shows different figures. It is not unusual to have different counts of houses from different sources. Part of the reason for a difference is that the various sources counting housing are looking at different time periods. But the Census also includes vacant housing units whole the other sources may not. Our experience in working in many other cities is that the Census numbers are likely too high due to containing empty and abandoned buildings

!

• Within that number are 9,050 residential single family homes. There are also 1,550

residential units that are part of multi-dwelling units (MDUs) meaning that they have 5 or more living units per building. The US Census showed almost 2,000 apartments in the city. But what we found is that there are 333 single ‘apartments’ or town-houses which are attached to single family homes, and that reconciles fairly well back to the Census number.

!

• In pricing out the network we have assumed that the cost for serving an MDU customer is

the same as serving a single family home. Sometimes the cost of serving MDUs can be lower. For examples, MDUs can often use fewer fiber pairs and can benefit by using larger ONTs. However, this savings can be offset by MDUs requiring more extensive wiring to make FTTH work. Our experience in other communities is that the cost of MDUs ends up being nearly as high as the cost of serving single family homes.

!

• Because the rural school addresses in the Austin Public School District are not

represented by a political entity, it was harder getting a count of the homes in the rural areas outside the city limits. We were able to get a count of land parcels that included an attached home from the Mower County Assessor’s office. We got something similar from Freeborn County. Those records gave us an estimate of rural residences as 1,767 in Mower County and 355 in Freeborn County.

!

• We also looked at several sources to estimate the number of businesses inside the city.

The 2010 Census set the number at 1,418. It has been our experience in working with other communities that the Census business counts are always generally not very accurate and can be either high or low. And so we looked at numerous other sources to get a count of businesses. We found the following. The Department of Revenue collects sales taxes from 544 businesses. The Austin Area Chamber of Commerce was told by a national source that there are 892 businesses in town that would be eligible to join the Chamber. A real estate web site says there are 1,106 businesses within a 2-mile radius of the center of Austin. And Austin Utilities estimates that there are 1,779 businesses in 1040 buildings. But we also have generally seen that utility business counts are a little high. This is due to the fact that the electric counts are based upon meters. For example, a business may have

a billboard in front for which the electricity is separately metered, and this might show as a business in the utility count. Businesses also sometimes request separate meters for each building they have or even for parts of a building, and if the corporate name for the two meters isn’t spelled identically in the records it will look like there are multiple businesses at the address. In our study we split decided to go with the real estate web site and used 1,100 businesses in the study. We used 150 rural businesses, outside the city limits in the study.

!

• The data on a PON fiber system is shared with up to 32 customers. The data isn’t literally

shared, but those customers have a shared pipe leading to the neighborhood. Some businesses do not want shared data and are willing to pay extra for data delivered only to them. The way to serve these types of businesses and the government is to design spare pairs of fiber in the network that can be used for ‘home-run’ fibers, meaning that a dedicated customer would get a fiber directly back to the headend, not shared with other customers or split by splitters. These businesses would also get a different type of termination equipment rather than an ONT, but at about the same cost.

!

• We have assumed the construction of a hut near to each of the existing electrical

substations. There would be some savings if we could collocate electronics inside the substations. We also have assumed one hut is needed north and one south of the city so that all customers are within twelve miles of a fiber hub, a requirement with PON technology. These huts house the splitters, which is where the fiber network is spliced to go from one fiber feeding a neighborhood to one fiber for every 32 households or businesses. This splitting is done physically and not electronically, which is why this is called a passive optical network.

!

• We also considered aerial fiber in the rural areas instead of burying the fiber.

Constructing aerial fiber means putting the fiber onto poles owned by somebody else. In the city, Austin Utilities owns the poles, making this an easy arrangement. But outside the city the poles are owned by other electric companies. In the city there is a significant savings from putting the network onto poles. The cost of burying fiber under city streets costs roughly twice as much per mile than putting it on poles. But the savings in the rural areas is more modest. We estimate that cable can be buried for about $27,000 per mile in the country or put onto poles for about $20,000 per mile. That savings would equate to a cost reduction of about $1.3 million. But there are other costs and considerations. First, the rules is that if you want to put a wire onto somebody else’s poles that you have to pay the full cost of doing so. This might involve rearranging the existing wires to make room for you or even in some cases might mean replacing poles that are already full with other wires. This is called ‘make ready’ in the industry and if the rural poles are typical then this could cost in the range of $300,000, reducing the savings. After that one must consider the long term benefits of having the cable buried. Buried cable lasts longer. It gets damaged less often. When considering all of this, it is not obvious that going with the lower net cost savings of $1 million is worth it for the project. The savings is there if

we need to lower the cost of the project, but if it can be afforded we think the long-term benefits of buried fiber make burying fiber in the rural areas the better choice.

!

Customer Electronics

• We priced the FTTH electronics based upon recent quotes we got from Calix. CCG is

vendor neutral and we are not suggesting that the community use Calix. Rather, our experience is that the cost of the FTTH electronics is similar between vendors and thus using a recent quote from any of the vendors is sufficient for predicting the cost of the network electronics. Calix just happened to be the most recent bid we have in hand.

!

• The ONT (Optic network terminal) at the premise is the device that converts light back

into the signals needed to provide the triple-play services. There are several different options for including ONTs in the study. First, ONTs can be external, meaning placed outside on the side of the building, or internal and placed inside like is done with cable modems. If the ONT is external, it has an optional battery that can keep the ONT running during a power failure. In the study we have modeled mostly internal ONTs. The upside is that inside ONTs are less expensive. They also can be plugged directly into an outlet for power, whereas outside a tap is needed into an external power wire. There is greater availability to technicians when the ONT is outside, but there is also the harsh Minnesota winter to deal with.

!

• Many FTTH networks have been designed with battery back-up in the ONT. However,

many of our clients have stopped providing batteries. The battery has historically been installed to operate phones in the case of a power outage at the home. However, there are fewer and fewer phones in existence that are powered from the phone line and most phones must be plugged into an outlet. So when such a phone loses power it can’t be powered by the battery. We have not included a battery in the design and would instead offer it as an option for a customer who really wanted it.

!

• Residential costs. The model assumes a total installed cost of a residential ONT at $453.

We further estimate the cost of a drop in the city at $502 including installation. This means it costs nearly $1,000 in incremental costs to add a residential customer to the network. This cost includes labor and materials for the fiber drop, installing the ONT, connecting to existing wiring, installing the settop box and instructing the customer about how to use the new system. The cost does not include the cost of settop boxes which some customers will get. In terms of total cash outlay, these are the costs when the installations are done by contractors. When installed by employees, the cost of paying the employee would be covered instead of an external installation labor costs. There is about $450 of labor included in the $1,000 cost. In rural areas the drops are expected to cost more and so the average rural drop will be about $50 higher.

!

• Business Costs. Small businesses can use the same ONT as residences and have the

rather than a lawn and it costs more to bury the drops. We have assumed that business drops in the city cost about $250 more than a residential drops. There will be some larger businesses that require a larger ONT. These large business ONTs which support many telephone lines plus a larger data connection cost $1,175 installed instead of $453. Large businesses sometimes also have more expensive drops since they tend to be the businesses with the really large parking lots in front that.

!

• In this model we have assumed the use of HPNA for distributing bandwidth within the

home. HPNA is a technology that uses the existing coaxial cable to deliver cable TV and broadband. There are other technologies available. For example, we also could have used MOCA which would use the existing coaxial cable in the house to deliver broadband. The cost for using various technologies for distributing the bandwidth throughout the home cost is roughly the same and the determination of the best technology can be determined in the future as needed.

!

• The ONTs we used in this model are designed to deliver only voice and data. This means

any Cable TV offering over the network must be digital and delivered over the data path. This requires the use of IPTV where video is 100% digital and delivered in an IP Data format to the set-top-box. IPTV is becoming the video delivery method of choice in the industry, so we used this technology in the model.

!

Rural Wireless Build

!

We also explored the alternative option of providing wireless broadband in the rural areas instead of fiber. Following are the assumptions used to make the estimates of the wireless cost.

!

• We assumed that the best wireless spectrum to use is 3.65 GHz. This spectrum can be

licensed at the FCC by anybody and currently is not used in Austin. I will provide a more detailed explanation of this spectrum later in the report.

!

• This spectrum will provide significant bandwidth on a point-to-point basis up to the

horizon. However, the bandwidth drops with distance from the transmitting tower, and so there is a practical limitation on how far a customer ought to be from a tower. We looked at two different version of a wireless network. There are maps showing these two options in Addendum I to the report.

!

• In the minimal network option we designed two towers. Customers within five miles of

these two towers would get speeds as fast as 28 Mbps close to the tower down to about 10 Mbps at 5 miles. But the speeds at the outer fringe of this network, which is about 9 miles would only get speeds of around 3 Mbps. While there are two existing water towers in town, they do not provide adequate coverage for the rural addresses in the Austin Public School District. So we proposed two new towers. One would be near Andytown, the other near the junction of Highways 21 and 28.

!

• We also looked at a second wireless design that instead uses five transmitters. Four of

these are at newly constructed towers and one of the transmitters goes onto an existing