Report S4
Natural Gas Movement Estimates - 1997
Table of Contents
1. Definition of the Commodity Flow Data Gap
1.1. General Description
Natural gas is composed primarily of methane but can also include ethane, propane, butane and pentane. In its purest form, such as the natural gas delivered to homes, it is almost pure methane. Methane is a molecule made up of one carbon atom and four hydrogen atoms, and is referred to as CH4.
Delivering natural gas is an enormous enterprise. Natural gas is transported to consumers through more than 300,000 miles of transmission pipelines with the help of vast storage reservoirs and thousands of compressors. This gas is sold to marketers, large commercial and industrial consumers, and distribution companies for delivery to consumers over a network of more than 1.1 million miles of local distribution pipelines.
Oil and gas extraction industries within NAICS code 211 are not covered by the Commodity Flow Survey (CFS), and thus, natural gas is out-of-scope.
1.2. Commodities Involved in the Data Gap
1.2.1. SCTG codes
Unlike other energy producing commodities (coal, crude petroleum, gasoline, fuel oil, and aviation turbine fuel), natural gas does not have a unique designated SCTG code. Instead, natural gas and liquefied natural gas (LNG) are classified with all other gaseous hydrocarbons in a gaseous state (excluding chemically pure) under the SCTG code 19330, or within liquefied gaseous hydrocarbons not elsewhere classified (excluding chemically pure) under the SCTG code 19329. That is, at the 2-digit level, natural gas is included in SCTG 19.
1.2.2. STCC codes
Under STCC coding, natural gas is included in code 13 - Crude Petroleum, Natural Gas, or Gasoline.
1.3. Establishments Involved in the Data Gap
1.3.1. NAICS codes
211 - Oil and Gas Extraction
486 - Pipeline Transportation
Both NAICS categories were excluded from the CFS 2002 sampling frame. This is an out-of-scope data gap for CFS.
1.3.2. NAICS-SIC conversion issues
There are no major NAICS-SIC conversion issues for this data gap. Natural gas is an out-of-scope commodity for the CFS.
2. Importance of the Data Gap
2.1. Value and Tonnage as a Share of National Shipments
Table 1 contains a national summary of Natural Gas movement estimates described in this report. Our estimates show that about 55 trillion cubic feet of inter-state Natural Gas were delivered in 1997. This is equivalent to about 1 billion tons shown in Table 1. In addition, we estimate non-local, intra-state movement of natural gas at about 22 trillion cubic feet, out of which about 11 trillion cubic feet were delivered to distributors for local distribution. The equivalent tonnages for these are also shown in Table 1. In total, we estimate total national Natural Gas deliveries at about 1.6 billion tons in 1997.
The natural gas movement estimates in Table 1 is much higher than the amount of natural gas consumed in 1997, estimated at 23 trillion cubic feet (or 435 million tons), due to transshipments among pipeline companies. Each time a shipment changes physical possession it is counted as a separate shipment in the EIA records. However, the ton-mile estimates appropriately include only the distance traveled at each stage of shipment as identified in Table 1.
2.2. Value and Tonnage as a Share for Individual Modes
Pipeline is the only mode of natural gas transportation, and therefore accounts for 100 percent of all movements.
2.3. Geographic Concentration: Dispersed versus Concentrated, Local versus Long Distance
Interstate movements as well as imports/exports of natural gas shipments are expected to be transported long-distances to various locations in the United States. In the absence of public data on average distance of natural gas movements we use the following numbers to calculate ton-miles. Based on average distance between state centroids we use an average length of inter-state movement of 240 miles. We use an average trip length of 100 miles and 30 miles, for the non-local and local natural gas movements, respectively.
2.4. Importance to International Trade
U.S. consumption of natural gas is increasing and is forecasted to continue. Natural gas demand is projected to grow by nearly 40 percent by 2025, to over 31 trillion cubic feet in 2025.1 The U.S. production of natural gas is significantly less than the total U.S. consumption. Historically, natural gas imports – mostly from Canada via pipeline – have made up this deficit.
Liquefied natural gas (LNG) imports represent an increasingly important part of the natural gas supply picture in the United States. LNG takes up much less space than gaseous natural gas, allowing it to be shipped much more efficiently from global production regions.
According to the EIA, the United States imported 3 trillion cubic feet and exported 0.7 trillion cubic feet of natural gas in 1997. Most of the current imports/exports are with Canada and Mexico via pipelines. Currently, only a small amount of the 3 trillion cubic feet of imports is in the form of LNG. LNG imports are expected to increase at an average annual rate of 15.8 percent, to levels of 4.80 trillion cubic feet of natural gas by 2025. LNG that is imported to the United States from countries other than Canada and Mexico comes via ocean tanker. The majority of U.S. LNG is from Trinidad and Tobago, Qatar, and Algeria. The United States also receives shipments from Nigeria, Oman, Australia, Indonesia, and the United Arab Emirates. On the exports side, Japan and Mexico are the two countries receiving LNG from the United States.
3. Data Sources
3.1. Coverage in CFS
The CFS does not include the natural gas industry; thus natural gas is out-of-scope. On the other hand, CFS does capture some of SCTG code 19 – Coal and petroleum products not elsewhere classified, so natural gas can also be treated as an under-count of in-scope CFS commodities.
3.2. Coverage in Other Data Sources
Natural gas movements, including imports and exports, are collected by EIA/DOE. Information related to interstate natural gas movements and international trade movements across U.S. borders is found in Table 8 of the 1997 Natural Gas Annual. Other related information such as natural gas points of entry and exit related to imports and exports, price information at different stages of the production process, and transportation, and the consumer sector (numerous tables) are found in the same publication. The EIA-176 database contains state level data collected from natural and synthetic gas producers, processors, distributors, storage operators, and pipeline operators. This database summarizes annual supply and disposition of Natural Gas by company in each state. In addition, a supplemental table contains state to state movements for the interstate portion of Natural Gas. The Federal Energy Regulatory Commission (FERC) also collects financial and operation data through FERC Forms 2 and 2A for major and non-major natural gas pipeline companies. The information collected is summarized by the Oil & Gas Journal and published annually in a special issue, Pipeline Economics, during August/September.
3.3. Data Quality
The tonnage and value of natural gas movements across state lines and U.S. border boundaries are based mainly on EIA published reports and databases. These data did not require any additional assumptions to arrive at national numbers. Thus, the data quality for the tonnage and value of natural gas moved across state lines and U.S. border boundaries is as good as those of the data source.
3.4. Other Issues
There are essentially three major types of natural gas pipelines along the transportation logistic: the gathering system, the transmission system, and the distribution system.
The gathering system consists of low pressure, small diameter pipelines that transport raw natural gas from the wellhead to the processing plant. The transmission natural gas pipeline network transports processed natural gas from processing plants in producing regions to those areas with high natural gas demands, particularly large, populated urban areas. Local distribution companies typically transport natural gas from delivery points (city gates) along transmission (interstate and intrastate) pipelines through thousands of miles of small-diameter distribution pipe and deliver it to individual end users.
Current estimates of natural gas tonnage, value, and ton-mile includes interstate, non-local intrastate and local intrastate portions of the transmission pipeline networks. The interstate portion of the transmission system, similar to the interstates highway network for truck freight, carries the major nationwide natural gas supplies to meet the demands of metropolitan areas. Therefore, the interstate portion is by far the largest.
4. Estimation Methods
4.1. General Description of Methods
We employed a combination of bottom-up and top-down approaches to estimate Natural Gas movements in the United States for 1997. Company data were taken directly from the EIA-176 Natural Gas database. These estimates were summarized to the state-level and disaggregated to FAF regions using data from the EIA on inter-state pipelines, and data from the Office of Pipeline Safety (OPS) on distribution pipelines.
4.2. Estimation of Regional Flows
We made use of the delivery records from the EIA-176 database for the estimation described in this report. Although the delivery records include those to end-use, inter-state, distribution companies, storage operators, own-uses and others, only three of these are included in our calculations. These are the inter-state movement, end-use and distribution company deliveries. The latter two are both intra-state movements. We adopted end-use delivery as non-local intra-state movements, which would include deliveries to large end-users and distributors. Deliveries to distribution companies are then taken as the part of non-local intrastate movement that is further distributed to local users.
4.2.1. Inter-State Natural Gas
Inter-state movements from the EIA-176 database were allocated to FAF regions using publicly available data on state-border capacity and average flow of transmission lines. This data was available for 1998, and included county of terminal for each pair of states. We identified associated FAF regions using a county to FAF region crosswalk. A number of the counties had city names rather than county names. For these records the corresponding counties were located using information from the GNIS (Geographic Names Information System). The capacity data was then summarized into a FAF to FAF region level for each state-level O-D combination and converted into a share matrix.2 This share matrix was then used in combination with the state to state Natural Gas flow matrix to calculate the FAF to FAF flows as:
INGk,m = Σi,j shi,k,j,m INGi,j
where i and j are origin/destination states, k and m are origin/destination FAF regions, INGi,j is state to state Natural Gas flow; INGk,m is FAF to FAF Natural Gas flow, and shi,k,j,m is the share matrix described above.
A few of the state to state flows were not captured in the above calculation because no pipeline records were present in the 1998 capacity data. FAF regions for these flows were assigned by comparing the indicated state borders on a map of FAF regions. Most of these had only one FAF origin/destination combination.
4.2.2. Local Intra-state Distribution
We use deliveries to local distribution companies from the EIA-176 as the state-level estimate of local intra-state movement of natural gas. We disaggregated these state-level records to the FAF regions using the following approach. First, we make the assumption that local distribution of natural gas is intra-county (and consequently intra-FAF), so that its FAF-to-FAF matrix would be diagonal. Second, we obtained data from the Office of Pipeline Safety (OPS) which contained information on company pipeline mileage for six pipeline categories (by diameter) at the city/county level. The mid-range of each diameter category was multiplied by its mileage to obtain a measure of distribution pipeline capacity for that category, which is then summed across all categories to obtain the company’s capacity. An estimate of the capacity in each FAF region of each state is then obtained by summing these numbers over the component counties. Dividing these latter numbers by the state total of company capacities provides an estimate of the relative capacity share of each FAF region in a given state. We use these shares to allocate the local distribution data to FAF regions as3:
LCNG j,m = shj,m × LCNGj
where j and m represent state and FAF regions, respectively, and LCNG is the local distribution of natural gas.
4.2.3. Non-Local Intra-state Distribution
Non-local, intra-state movements are estimated from the EIA-176 database as the sum of deliveries to distributors, deliveries to certain end-users not expected to be served by local distributors (e.g. electric plants and vehicle fuel), and any end-use deliveries of natural gas not owned by the delivering company. By definition these movements occur between state-borders and city gates. We disaggregate the movements to FAF regions as follows:
NLNG j,k,m = shj,k × shj,m × NLNGj
where j, k, and m has the same meanings as above, and NLNG stands for non-local, intra-state natural gas movements. The parameters shj,k and shj,m are vectors allocating each states non-local natural gas movements to its component origin and destination FAF regions, respectively. shj,k is calculated from the FAF destination sum of inter-state natural gas flow for each state, while shj,m is the same vector used to allocate local gas distribution in 4.2.2 above.
4.3. Expected Quality of the Estimates
Estimated FAF regional natural gas movements will be based on state-level information published by EIA. Thus, the quality of the FAF regional flows is inherited from EIA's state-level published information.
5. Implications for the Scope and Content of the 2007 CFS
Since state-level natural gas movements are collected by the EIA on a continuous basis, no duplication of effort by the CFS is necessary. However, given the assumptions employed in disaggregating these data to FAF regions better accuracy suggests a need to construct more realistic pipeline network tables for allocating state level data to FAF regions.
6. References
- U.S. DOE (1997) “Annual Report of Natural and Supplemental Gas Supply & Disposition” Form EIA-176.
- U.S. DOT (1997) 1997 Distribution Annuals Data http://ops.dot.gov/stats/DT98.htm .
- U.S. DOE (1998) “Deliverability on the Interstate Natural Gas Pipeline System” DOE/EIA-0618(98).
1 Annual Energy Outlook 2005 with
Projections to 2025, DOE/EIA-0383(2005), January, 2005. http://www.eia.doe.gov/oiaf/aeo/index.html
2 We adopted pipeline capacity data
instead of flow rates for two reasons. One, average flow rates were
zero for some pipelines, probably because these are newer pipelines.
Two, many of the pipelines display a similar capacity to average
flow rate ratio.
3 Note that this approach, borne out of
lack of any other publicly available data, rests on two assumptions
that we consider reasonable. One, we assume that pipeline mileages
associated with company county in the OPS data are located in the same
county. This assumption matches with the “city” gate boundary
of the local distribution data. In addition, since the estimates are at
a much higher level of aggregation (FAF) errors due to misplacement of
pipeline locations are minimized. Eighteen of the states have only one
FAF region, and only 7 have more than 3 FAF regions. Two, we assume
that the calculated capacity shares represent the allocation of distribution
to FAF regions in each state.
