1、 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases API Statistics Department January 11, 2018 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases API Statistics Department January 11, 2018 COPYRIGHT NOTICE All information offered in this report is the sole and exclusive prope
2、rty of the American Petroleum Institute. You may not reproduce, upload, post, transmit, download, or distribute, resell or otherwise transfer outside of your company without express consent of the American Petroleum Institute. 2018 American Petroleum Institute. All rights reserved. 2016 Sales of Nat
3、ural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved. iii Protection of API Data WARNING LIMITATION OF LIABILITY: This data is based on information that has been voluntarily reported to the American Petroleum Institute by petroleum companies operat
4、ing in the U.S. The American Petroleum Institute does not guarantee the accuracy of the data and disclaims any express or implied liability in connection with its use. COPYRIGHT: The American Petroleum Institute used considerable expertise and judgment in compiling the statistics that were used in t
5、his report. The report and all of the statistical content contained in this report are the sole and exclusive property of the American Petroleum Institute and are specifically protected by the Copyright Act (17 U. S. C. 101). These statistics may not be copied, downloaded, uploaded, published, store
6、d, distributed, or transmitted in any manner without the express written authorization of the American Petroleum Institute. TRADE SECRET: The data and statistical information contained in this report are valuable trade secrets owned by the American Petroleum Institute. The use, copying, downloading,
7、 uploading, transmitting, sending, communicating, conveying, publishing, or disclosure of any of this information without the express written authorization of American Petroleum Institute is prohibited by D. C. Code Ann. 48-501 (1989), state statutes in most jurisdictions, and by Federal law pursuan
8、t to the Economic Espionage Act of 1996 (18 U. S. C. 1831). 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved. v Table of Contents Introduction vii Summary ix Tables A. Sales of Natural Gas Liquids and Liquefied Refinery Gases b
9、y PAD District and Product Type: 2016 and 2015 . 1 B. Sales of Propane in the United States, by End Use, PAD District, and State: 2016 and 2015 2 C. Sales of Odorized Propane in the United Sates, by End Use, PAD District, and State: 2016 and 2015 . 4 D. Five-Year History of State Sales and Proportio
10、n of Odorized Propane, in the United States . 6 E. Five-Year History of Total Sales of Odorized Propane Versus Total Sales of Propane, in the United States by PAD District and State . 7 F. Sales of Odorized Propane per Customer Account, in the United States, by End Use, PAD District, and State: 2016
11、 and 2015 . 8 Appendices A. Estimation Methodology 9 B. Data Collection Form: 2016 Sales of Propane Report 14 C. Instructions and Definitions: 2016 Sales of Propane Report . 16 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved.
12、v 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved. vii Introduction The 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases presents the results of the thirty-second annual survey jointly sponsored by the American P
13、etroleum Institute (API), GPA Midstream Association, National Propane Gas Association (NPGA), and Propane Education tertiary storage is defined to be all inventory held in customer tanks. All propane held in retailer storage tanks or customer storage tanks will ultimately be consumed. Furthermore, a
14、ll volumes of propane in retailer and customer storage tanks (production, primary inventory, imports) are captured by EIAs data collection system and will not be returned (in any significant volumes) into the primary supply distribution system. However, a few petrochemical companies have private sto
15、rage facilities. These facilities are generally underground salt cavern storage and are connected via pipeline to primary storage facilities, such as those in Mont Belvieu. Propane stored in private salt caverns may be returned to primary storage. When inventory from private salt cavern storage is r
16、eturned to the primary distribution system and immediately sold into the market, the EIA data collection system will not capture these volumes in its annual summary of propane supply. In 2016, PERC receipts, raw survey data and other analyses indicated substantial volumes of uncounted supply were de
17、livered into the primary distribution system and were immediately redelivered into the market. The results of the 2016 propane sales survey include an adjustment to U.S. propane supply statistics published by EIA in the Petroleum Supply Annual for 2016. 2016 Sales of Natural Gas Liquids and Liquefie
18、d Refinery Gases 2018 The American Petroleum Institute. All rights reserved. ix Summary In 2016, sales of natural gas liquids and liquefied refinery gas totaled 44.5 billion gallons. This represents a decrease of 8.8% from 2015. Sales of Natural Gas Liquids and Liquefied Refinery Gases: 2016 and 201
19、5 (Thousands of Gallons) Volume Percent2016 2015 Change ChangeButane19,162,384 9,049,152 113,232 1.3%Ethane117,623,677 16,941,360 682,317 4.0%Pentanes Plus13,588,480 4,025,658 -437,178 -10.9%Propane 14,152,824 18,831,317 -4,678,493 -24.8%United States Total 44,527,365 48,847,487 -4,320,122 -8.8%Prod
20、uct Type1Petroleum Supply Annual 2016 Volume 1, Tables 3, 5, 7, 9, 11 and Petral Consulting CompanyButane20.6%Ethane39.6%Pentanes Plus8.1%Propane31.8%2016 Sales by Product Type2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved. x
21、 The tables below show the states with the largest odorized propane sales in the residential, commercial, and agricultural sectors. Totals may not equal the sum of their components due to independent rounding. Rank State1. Michigan 360,829 8.3%2. Wisconsin 237,607 5.4%3. California 220,511 5.1%4. Ne
22、w York 219,679 5.0%5. Minnesota 212,104 4.9%Total 1,250,730 28.7%Rank State1. California 119,950 7.3%2. Florida 97,402 5.9%3. Pennsylvania 88,953 5.4%4. New York 86,543 5.3%5. Virginia 79,716 4.9%Total 472,564 28.8%Rank State1. Iowa 142,697 16.9%2. Minnesota 87,683 10.4%3. North Carolina 67,074 7.9%
23、4. California 52,126 6.2%5. Illinois 44,642 5.3%Total 394,223 46.7%Residential SectorVolumes (Thousands of Gallons)Percent of Total ResidentialCommercial SectorVolumes (Thousands of Gallons)Percent of Total CommercialAgricultural SectorVolumes (Thousands of Gallons)Percent of Total Agricultural2016
24、Sales of Natural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved. 1 Table A.2016 2015 2016 2015 2016 2015 2016 2015PADD 1 974,862 781,200 0 0 536,424 457,044 3,624,516 3,274,846PADD 2 1,657,782 1,818,264 947,058 1,037,994 767,802 628,236 3,596,334
25、4,825,684PADD 3 5,814,732 5,965,932 16,571,619 15,830,832 1,742,916 2,400,174 5,392,506 9,417,216PADD 4 117,432 153,720 104,160 71,064 129,486 102,732 610,764 593,514PADD 5 597,576 330,036 840 1,470 411,852 437,472 928,704 720,057U.S. TOTAL 9,162,384 9,049,152 17,623,677 16,941,360 3,588,480 4,025,6
26、58 14,152,824 18,831,3171Petroleum Supply Annual 2016 Volume 1, Tables 3, 5, 7, 9, 11 and Petral Consulting CompanySales of Natural Gas Liquids and Liquefied Refinery Gases by PAD District and Product Type: 2016 and 2015(Thousands fo Gallons)PAD DistrictButane1Ethane1Pentanes Plus1Propane2016 Sales
27、of Natural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved. 7 Table B. Sales of Propane in the United States, by End Use, PAD District, and State: 2016 and 2015District API refers to them as multi-state retailers 2. All other propane retail markete
28、rs; API refers to them as single-state retailers but understands that they may operate in more than one state. The response rate for multi-state retailers is considered to be 100%, therefore no adjustment or estimation is needed. The estimation methodology therefore centers on single-state retailers
29、. API estimates the Propane sales for the single-state retailers then appends the volumes to the reported retail propane sales gallons from large retailers to arrive at the estimated total propane sales for each state. Therefore, all further discussion will center on the estimation methodology used
30、to calculate the single-state data and will ignore the multi-state data. II. Estimation Methodology Step 1 Collecting Publicly Available Data Retail sales of propane data for total U.S., by PADD, and, whenever possible, by end-use are collected from publicly available sources. Much of the data on th
31、e PADD level (and even some state detail) are available publicly from Energy Information Administration (EIA) and Petral Consulting Company. Depending on the time of the year, API can obtain the following data: 1. Total Propane Sales by PADD 2. Propane sales to Chemical plants by State These data ar
32、e therefore with one minor adjustment - not estimated by API but used to calculate the sales by end use for each state. Historically, propane import statistics published by EIA and the Foreign Trade Division/Census Bureau were in close agreement. Furthermore, propane import statistics published by t
33、he National Energy Board of Canada were also in close agreement with Foreign Trade Division/Census Bureau statistics. In 1993, however, EIA statistics for propane imports were 39.1 M Bpd less than Foreign Trade Division statistics. Since 1993, EIAs published data for propane imports have been chroni
34、cally less propane imports as published by the Foreign Trade Division/Census Bureau. The understatement of imports is typically in the range of 25-50 M Bpd. Based on a detailed review of PADD level propane import statistics, EIA statistics consistently understate Canadian propane imports into PADD 1
35、 and PADD 2. The understatement of propane imports causes EIA statistics to understate total propane supply especially in PADD 1 and PADD 2. EIA propane supply statistics are critical to the preparation of the annual report of propane sales by end-use category. In particular, EIAs undercount of prop
36、ane imports would cause propane sales into the residential/commercial markets to be understated if total propane supply statistics were not adjusted to reflect this chronic undercount problem. Hence, EIAs published propane supply statistics are adjusted to reflect the understatement of Canadian prop
37、ane imports. The adjustment to EIA propane supply statistics improves the quality of the propane end-use sales presented in the annual Sales of Natural Gas Liquids and Liquefied Refinery Gases. In 2004, concern about EIAs accurate measuring of PADD to PADD transfers of propane were heightened when r
38、eported data exceeded the EIA PADD total for the first time. It was therefore determined that better overall estimates could be achieved if regional PADD controls were lifted and the total U.S. number was used. In essence to keep the nomenclature of this methodology unchanged all the states were con
39、sidered to belong in one PADD and the estimated states totals were rolled up directly to the U.S. totals, not the PADD totals.APPENDIX A 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases 2018 American Petroleum Institute. All rights reserved. 10 Step 2 Adjusting the Reported Numbers to
40、Reflect Response Rate Response will vary state. While API may receive responses from 10 of 15 retailers from Alabama, there may only be 2 of 20 responses from Virginia. In general, response rates for this survey are between 40 and 95 percent within a state these rates are sufficient to give API a hi
41、gh confidence in its estimates. Assuming that response within a state is independent of volume (a detailed discussion of this assumption can be found in section III. Methodology assumptions and caveats), we can adjust the total reported numbers within each state as follows: Adjusted total reported p
42、ropane sales by state = Tj =jjnccnNP =1Where, P = Total reported retail propane sales volumes per company n = Number of respondents in this state N = Number of surveys sent to this state c = A responding company j = A state Step 3 Estimating Total Retail Propane Sales by State At this point, we have
43、 two estimates for each PADD: 1. The publicly available PADD total (from EIA) 2. The estimated PADD total reached by adding the estimated state totals calculated in step 2 above. The difference between the two totals is simply the normal variance one observes, as an estimate is calculated using two
44、different estimation methodologies. API has decided that the publicly available numbers, unless the difference between the two numbers is substantial1, are the ones to use in this publication. We must therefore adjust the estimated state totals calculated in step 2 above to add-up to the publicly av
45、ailable total for each PADD. Again, we use proportional allocation to reach our estimates as follows: Estimated total propane sales by state = Ej = knjjkjkDTT=1Where, k = PADD number (k = 1 to 5) Tj = Adjusted total reported Propane sales by state (see step 2 above) D = Total Estimated Propane sales
46、 in that PADD j = A State n = Total number of states in that PADD 1The definition of the word “substantial”, which is also mentioned in a later section of this methodology, is left to the statistician working on the problem to define (based on the relative data in question). Another way to explain i
47、t is by using the phrase “statistically significant”. APPENDIX A 2016 Sales of Natural Gas Liquids and Liquefied Refinery Gases 2018 The American Petroleum Institute. All rights reserved. 11 Steps 2 and 3 - Example In plain English, the idea behind step 2 is simple: since the response rate varies fr
48、om state to state, we first adjust the reported propane numbers to reflect a 100% response. Therefore, if State A reported 1 million gallons with a response rate of 20%; that translates into an estimated 5 million gallons had we received a 100% response (1 million 0.2 = 5 million). Moreover, if Stat
49、e B also reported 1 million gallons but with a response rate of 50%, this translates into an estimated reported total of 2 million gallons with a 100% response (1 million 0.5 = 2 million). Assuming that this PADD consists of only these two states, the PADD total is therefore 5 million + 2 million = 7 million gallons. To explain step 3, we need to then look at the EIA PADD totals and adjust our state estimates so they add-up to that known total. Assuming that the EIA total for the PADD is actually 7.1 million gallons, we can
