Modinput

RTF - FCC Model InputsOverview: 

RTF - FCC Model Inputs Overview Presented by Bob Schoonmaker

Overview: 

Overview Customer Location Outside plant - general Cable cost & construction Structure Sharing Serving Area Interface & Digital Loop Carrier (DLC) Switching Cost Expenses

Objective: 

Objective Discuss current status of FCC non-rural inputs Review major sources of non-rural data Discuss what additional review and analysis should be done

Customer Location: 

Customer Location Customer Locations - Road surrogate Access lines Study area lines => ARMIS Adjusted up to ARMIS if < study area lines Calculated to CBG - disaggregated to CB Plant built to occupied vs. all households Accuracy of BLR wire center boundaries - comments requested

Outside Plant - General: 

Outside Plant - General Assumes cable costs for feeder and dist. are identical Uses 26 gauge and 24 gauge copper based on a copper cross-over point - 12,000 ft. Fiber to all loops that exceed 18,000’ or are “cost-effective”

Engineering Assumptions: 

Engineering Assumptions Model optimization “on” for clusters less than 500 lines (significant impact on run time) T-1 technology between 18,000 and 24,000 ft. not used Road factor - model uses rectilinear distances so no additional distance multiplier to reflect road vs. model distances

Cable Cost General: 

24 and 26 gauge costs developed for all cable sizes Separate cost adopted for aerial, buried and underground to reflect construction cost Separate cost installation schedules for normal, soft rock, and hard rock Cable Cost General

NRRI Study: 

NRRI Study David Gabel study of RUS data Work orders for three years Developed cost for cable and cable construction Used as the basis for much of the outside plant construction costs Analysis used a least squares regression technique

Cable Cost - Copper: 

Cable Cost - Copper FCC modified regression to a Huber rather than least squares regression - tends to eliminate outlier values Regression formula has fixed cost/foot plus cost per pair plus factor for more than one cable at a location. Separate cost analysis for cable structures

Cable Cost - Copper (Cont’d): 

Cable Cost - Copper (Cont’d) RUS data analysis with 3 adjustments - 24 gauge aerial Non-rural LEC superior buying power - 15.2% reduction for aerial based on Bell Atlantic to RUS 10% loading for LEC engineering (but not contractor engineering which is in RUS #s) 9.4% for splicing (not in Gabel analysis)

Cable Cost - Copper (Cont’d): 

Cable Cost - Copper (Cont’d) 24 gauge underground Same 3 adjustments but 16.3% for buying power 24 gauge buried Same 3 adjustments with 15.2% for buying power 26 gauge cable Ratio to 24 gauge based on regression analysis of Aliant, Sprint and BCPM default values

Cable Cost - Fiber: 

Cable Cost - Fiber NRRI modified for Huber regression-3 adjustments Aerial Fiber - 33.8% reduction for buying power - 10% LEC engineering - splicing 4.7% Underground Fiber - 3 adj. - 27.8% for buying power Buried Fiber - 3 adj. - 27.8% for buying power

Cable Fill Factors: 

Cable Fill Factors Should reflect “current demand” not “ultimate demand” Factor inputs reflect engineering points, not actual fill Select HAI defaults for distribution -50% to 75%

Cable Fill Factors (cont.): 

Cable Fill Factors (cont.) Feeder Fill Factors - Average of HAI and BCPM Vary from 70% to 82.5% Fiber Fill Factors - 100% since redundancy is built into 4 fibers for multiplexer

Structure Costs - General: 

Structure Costs - General Model cost input by: Structure type Distribution or feeder Normal, soft rock, hard rock

Structure Costs - General (Cont’d): 

Structure Costs - General (Cont’d) Aerial structure NRRI - 40’ Class 4 material - 213.94 NRRI - anchors, guys & other material - 32.98 rural, 49.96 suburban, 60.47 urban LEC engineering - 10% of pole and guys Converted into costs/foot for density zones, zone 1-2 rural, 3-7 suburban, 8-9 urban Pole distance Z1-2-250’, Z3-4-200’, Z5-6-175’, Z7-9-150’ Generally consistent with BCPM and HAI

Structure Costs - General (Cont’d): 

Underground Structure Subsequent study of RUS data by NRRI Cost per foot, variability for bedrock and soil type indicators 10% for LEC engineering Extrapolate zones 3-9 based on ratios of BCPM and HAI input values Structure Costs - General (Cont’d)

Structure Costs - General (Cont’d): 

Cost of buried structure 24 gauge copper formula is for both materials and structure Modified formula - to identify cost of the structure only Extrapolate density zones 3-9 using same method as underground Structure Costs - General (Cont’d)

Plant Mix: 

Plant Mix Three tables - distribution, copper feeder, fiber feeder Distribution Plant - Plant Mix Underground - 0% to 90% Buried - 60% to 0% Aerial - 40% to 10%

Plant Mix (cont.): 

Plant Mix (cont.) Feeder Plant - Density Underground - 5 to 95% Buried - 50 to 0% Aerial - 45 to 5% For both copper and fiber

Structure Sharing: 

Structure Sharing % Used by telco Aerial - 50% zones 1-6, 35% zones 7-9 UG & Buried - 90% in 1-2 85% in 3 65% in 4-6 55% in 7-9

SA Interfaces: 

SA Interfaces To connect copper feeder and distribution 18 SAI sizes from 7200 to 1 line Extrapolate cost other than 7200 line from 7200 pair indoor - largest cost is splicing and protectors Take HAI cost for other sizes times 2.25

Digital Loop Carrier: 

Digital Loop Carrier 5 different DLC sizes Based on cost data submitted by some telcos Reduced by 2.6%/year for cost reduction in switch costs

Switching Cost: 

Switching Cost Based on data from FCC depreciation filings from 20 states and 139 observations (921 switches) From RUS data - 1995-6 - 139 switches Switches removed RBOC under 1,000 lines “Outlier” switches - high or low line cost RUS Upgrades Switches over three years old

Switching Cost: 

Switching Cost Adjustments to RUS data for costs not included: RUS didn’t include: MDF cost - $12/line Power 0-999 $12,000; 1,000-4,999 $40,000; 5,000-25,000 $74,500 Terminating Remote - $27,598 per Gabel Engineering - 8% Gabel estimate

Switching Cost: 

Switching Cost Doesn’t include switch upgrade costs Purchasing and installation increased by inflation Switch cost - time in regression equation

Switching Cost: 

Switching Cost Regression results Fixed Remote - $186,000 Fixed Host - $447,000 Per Line-all - $83 HAI default uses a blended switch cost configuration FCC proposes to use LERG to identify host/remote

Switching Cost-Other: 

Switching Cost-Other Max size - 80,000 lines Switch port fill - 94% Trunking Model computation on 6:1 trunk ratio should be disabled Trunk cost for tandem switch - $100/line

Plant Specific Expenses: 

Plant Specific Expenses Single national factor for each plant type % of investment Current to book investment ratios based on replacement cost - 4 RBOCS & GTE Expense to “current” investment ratio Adjusted forward from 1996 by 6% productivity reduction plus inflation increase

Support Expenses: 

Support Expenses Cust. Ops., Corp. Ops, Network Ops., Other non-plant specific Calculated on a per line basis Based on 1996 ARMIS data Regression analysis of lines and toll minutes to determine what is applicable to USF Time adjusted by 6% productivity reduction plus inflation increase

Customer Operations Expense: 

Customer Operations Expense Only ETC required Marketing expense included - 95.6% excluded Adjustments to exclude activities for non-basic services Coin collection Directory publishing Carrier access billing (CABS) Carrier customer contact Service Order Processing

Customer Operations Expense (cont.): 

Customer Operations Expense (cont.) After adjustment and regression - 28.7% assigned to USF Cost/line calculated Adjusted to forward looking using 6% productivity reduction plus inflation increase Proposed per line amount - $1.07 Customer Service, $0.02 Marketing

Corporate Operations Expense: 

Corporate Operations Expense Regression analysis adjustment only 60.0% of actual assigned to USF Proposed per line amount - $2.60

Network Operations & Other Non-Plant Specific: 

Network Operations & Other Non-Plant Specific Network Operations - 48.7% assigned to USF Proposed per line amount - $1.35 Other non-plant specific - 385.5% of actual assigned to USF Proposed per line amount - $0.07

Where do we go from here?: 

Where do we go from here? Review data after final FCC order adopted Redo plant construction cost without volume purchase discounts Huber regressions vs. least squares (?) Analyze RUS only switching data (?) Eliminate productivity adjustments from any expense adjustments

Where do we go from here? (cont.): 

Where do we go from here? (cont.) Uncertain about other plant specific adjustments Definitely have issues with cost/line development for other expenses (customer ops., corporate ops., etc.) Local/toll regression analysis on expense (?) Need a source for customer operations expense

Where do we go from here? (cont.): 

Where do we go from here? (cont.) Structure sharing assumption changes and support Plant mix assumptions - How do we account for variations? How do we deal with extreme outlier situations? Do we want to try to gather data to validate individual cost items?

VALIDATION: 

VALIDATION

Evaluations of FCC Model: 

Evaluations of FCC Model Model Design Documentation review Review of comments Analysis by outside parties, as available

Evaluations of FCC Model: 

Evaluations of FCC Model Input Evaluation - Prioritization and variability of inputs Sensitivity analysis Judgmental evaluation Importance to rural telco results Perceived ability to influence - Input Values - Review of FCC orders and comments - Other evaluations as appropriate

Evaluation of FCC Model: 

Evaluation of FCC Model Evaluation of model results Comparisons to Actual - Physical units - Access lines - Sheath miles of plant - Network location (NECA mapping analysis) - Clustering - # of clusters - Investment $ - Expense $

Comparisons to Actual: 

Comparisons to Actual Validity varies depending on the specific investment or expense item Model network may be different architecture from actual network Model reflects current construction costs, actual investment includes historical costs Model only develops cost of “local” service not total company costs

Comparisons to Actual (cont.): 

Comparisons to Actual (cont.) Plant investments COE Switching - Personal opinion - Should be a reasonable match COE Transmission - Personal opinion - Reasonable match if there is recently built DLC plant Outside Plant - Personal opinion -Reasonable match if company has recently built DLC plant Support Plant - Should have some relationship to actual

Comparisons to Actual (cont.): 

Comparisons to Actual (cont.) Expenses COE Switching - Should be a reasonable match Other plant specific - May differ because of investment differences Customer Operations - Actual needs to be adjusted to remove “Non-USF” portion in order to compare Corporate Operations - Same as customer operations

Current Validation Approach: 

Current Validation Approach Approximately 20 companies representing rural company characteristics - All regions of the country - All sizes of rural companies - Varying company types (coop, stockholder, holding company) - Some companies with actual DLC network

Regions of the Country: 

Regions of the Country a. New England b. Other Northeast c. Southeast d. Upper midwest e. Lower midwest f. Southwest g. Mountain States h. Northwest i. Alaska j. Insular

Company sizes: 

Company sizes Very Small- Under 500 lines Small – 500 – 2000 lines Small medium 2000 – 5000 lines Medium – 5000 – 10,000 lines Medium/Large 10,000 to 50,000 lines Large – Over 50,000 lines

Current Validation Approach: 

Current Validation Approach Initial limit on companies studied to limit resource requirements Review model results for various components of network costs “What if” analysis of some inputs Intent is to identify issues, concerns, validity of results

Current Validation Approach: 

Current Validation Approach Possible testing of certain results on wider groups of companies as data is available Analysis of total results for all rural companies on limited number of input sets Develop report and recommendations