LDC SUPPLY PORTFOLIO MANAGEMENT DURING THE WINTER HEATING SEASON

Transcription

1 EA December 12, 2013 LDC SUPPLY PORTFOLIO MANAGEMENT DURING THE WINTER HEATING SEASON I. Introduction Each year local natural gas utilities develop a plan to reliably meet customer needs during winter heating season peak consumption periods. The plan is usually based on a forecast of expected loads and is later adjusted to actual weather-induced demand requirements. Numerous scenarios are examined when building a seasonal natural gas supply portfolio always against the backdrop of normal, which is defined by companies based on local weather information and system requirements from years past. Supply tools, such as firm pipeline capacity, access to on-system or pipeline storage, peak-shaving capabilities, local production and even third-party transportation arrangements, are carefully considered. Plans to manage supply pricing risks may also be in place. In many cases, these plans are submitted to state regulators for approval prior to the start of the winter heating season. As local gas utilities and natural gas consumers approached the winter heating season (which is the period of November 2012 through March 2013), market acquisition prices had been primarily falling since August In fact February 2012 pricing at Henry Hub actually fell to below $3.00 per MMBtu and remained that way from February through October Since then winter prices on a national average basis grew slightly too more than 4.00 per MMBtu but fell again in the spring and summer of Remarkably even with lower wellhead prices, domestic natural gas production grew about five percent in 2012 over 2011 and grew another two percent to reach record highs of 67 Bcf per day by year-end 2013 (according to Bentek Energy, LLC). Rounding out a positive gas supply story for the period leading up to the winter heating season, underground storage working gas volumes in 2012 began the April-October net injection season at about 2.5 Tcf, a very high volume compared to history, which meant summer injections of working gas to reach a pre-winter full underground storage would be moderate compared to prior requirements. As it turned out however, falling natural gas prices in 2012 influenced demand for natural gas into power generation (primarily displacing coal-fired generation), ultimately increasing gas consumption in the power sector by 20 percent over 2011 an extraordinary short-term increase. Again remarkably, the incremental power demand for natural gas had little upward influence on market prices during Simply put; natural gas that may have previously gone to summer working gas injections went instead to generating electricity with little apparent pricing impact by the American Gas Association

2 Given this backdrop, the analysis described in this paper regarding critical elements of the winter heating season (WHS) originates from data acquired from AGA member local distribution companies (LDCs) through the AGA LDC Winter Heating Season Performance Survey. For this year s survey, questions focused on peak-day and peak-month supply practices, pricing mechanisms, as well as regulatory and market hedging practices. This year responses (whole or subsets) were received from 74 local gas utilities with service territories in 44 states. The sample companies had an aggregate peak-day sendout of 54.1 million Dekatherm (Dth), acknowledging that the peak-day did not occur on the same calendar day for each company. However, these same companies planned for a peak-day of 71.8 million Dth in aggregate, which means that only about 75 percent of the planned peak sendout volume was actually required during the WHS. This makes the tenth consecutive year that aggregate actual peak-day sendout fell short of aggregate design peak-day volumes for responding companies. The purpose of this report is to document gas delivery system operations of the surveyed local gas utilities during the past winter heating season and to help provide insights into gas supply trends and procurement portfolio management. The aggregated data presented in this report are not to be interpreted as standards or best practices for gas supply management. Instead they represent a snapshot of aggregated supply procurement practices of those companies that participated in this year s survey. The need to implement and the timing of any implementation of highlighted practices will vary with each operator based on, among other things, their unique regulatory, geographic and operational characteristics. In some cases, the report compares survey results for the winter heating season with those reported one year or several years prior. It should be noted, however, that the compared samples are not identical and the supporting data are not audited or normalized for sample differences, weather or other factors. II. Executive Summary This report is based on survey responses submitted by 74 AGA member local gas utilities. These companies had a cumulative, non-coincident, peak-day sendout of 54.1 million Dth and an average peak-day sendout of 730,781 Dth, which is about the same as last year s sample ( winter) of 732,970 Dth. The coldest day, as reported by respondents, occurred predominantly in January (70 of 74 respondents), which is similar to the winter heating season (with 53 of 63 respondents reporting a January peak day). Results in this winter heating season survey are generally presented as counts of companies that fit into percentage ranges of supply volumes (e.g., 1-25%, 26-50%, and so forth). The intent of this report is to document the data as a snapshot of supply behavior by large purchasers of natural gas in this case the surveyed local distribution companies (LDCs). Natural Gas Market The U.S. natural gas market balances supply and consumption today at above 70 Bcf per day on average. However, requirements for natural gas by consumers and particularly during the winter heating season are not average. During the period of November 1, 2012 through March 31, 2013, total consumption of natural gas in the U.S. ranged from about 70 Bcf per day on a warm March day to over 110 Bcf (including net exports to Mexico) on the coldest winter day in January (according to Bentek Energy LLC) a huge swing in observed daily winter heating season demand. 2

3 In fact the residential and commercial sectors of the market were most responsible for the dramatic swings in customer requirements, ranging from over 60 Bcf per day on a cold January day in 2013 to a winter heating season low of 25 Bcf per day in late March. Weather For the two months (September-October) just prior to winter heating season, conditions were 4.6 and 3.2 percent warmer than normal, respectively. This pattern continued into the winter with warmer than normal temperatures for the nation as a whole recorded during the months of November-December 2012 and January Only February 2013 (2.0 percent colder) and March 2013 (10.5 percent colder than normal) deviated from the trend. In fact the National Oceanographic Atmospheric Administration reported that the first six months of 2012 were the warmest ever recorded for a January through June period in the lower-48 states. That trend sustained itself for the balance of the year only finally turning colder in February-March Even though February and March were colder than normal, the peak consumption day occurred in January for 70 of the 74 surveyed companies, while two identified December and two selected February or March as the month in which their peak day load occurred. Temperatures around the country were consistently warmer than normal throughout the early winter heating season months and colder on average during the later stages of the winter. For the period of October 1, 2012 through March 31, 2013, cumulative heating degree days were 3.5 percent fewer than normal on a national basis (meaning warmer than normal), while the previous winter season ( ) in contrast had been 17.5 percent warmer than normal for the nation as a whole. On a regional basis for the winter, cumulative temperatures were warmer than normal in every region, ranging from 1.7 percent warmer in the West North Central area to 10.1 percent warmer in the West South Central region. For an example of cumulative weather resulting in a consistently colder-than-normal winter, one must look as far back as Winter weather has been decidedly warmer than normal on average compared to the 30-year norm since that remarkable winter, when sustained cold temperatures and concerns regarding a tight supply market resulted in significant natural gas price leaps. Gas Supply Portfolios Local gas utilities build and manage a portfolio of supply, storage and transportation services, which include a diverse set of contractual and pricing arrangements, to meet anticipated peak-day and peak-month gas requirements. For the winter heating season, companies responding to the AGA survey planned for 71.8 million Dth of peak-day gas sendout, but only 75 percent (54.1 million Dth) of the volume was actually required because of the lower than projected peak consumption levels nationwide. As a point of reference, last year s sample of 63 companies planned to deliver about 66.8 million Dth of peak-day gas requirements but in fact delivered only about 46.2 million Dth (about 69 percent). Local gas utilities apply a standard or methodology for determining a design peak day temperature calculation and of course that influences the construct of their gas supply portfolio. For the WHS survey, nineteen companies noted using a 1-in-30 year risk or probability of occurrence, while 26 companies choose other time periods. Twenty-eight companies used other methodologies including a historical peak (e.g., coldest recorded temperature since 1970), Monte Carlo statistical simulation, and coldest effective degree day in a 30-year period. 3

4 It should be no surprise that purchases moved by firm transportation provided much of the gas to consumers for the peak day and peak month. Sixty-two of 74 companies indicated that firm supplies were a part of their gas supply portfolio, including 43 companies that used firm supplies to meet between 26 and 75 percent of their peak-day volume requirements. Fifty-seven companies indicated that up to 50 percent of peak-day supplies originated from pipeline or other storage; 51 companies noted that up to 50 percent of the deliveries arriving at their city gate on a peak day were earmarked for transportation customers on their system; and 16 companies flagged on-system storage as the source of up to 50 percent of peak-day supplies. Mid-term (more than one month, less than one year) agreements were the most utilized for peak-day purchases, with 62 of 72 companies having such contract terms. Moreover, 30 companies indicated that more than 50 percent of their peak-day natural gas supplies were acquired via mid-term agreements. Long-term agreements, defined as one year or longer, were used by 39 of 72 reporting companies within their peak-day gas supply portfolio (compared to 36 of 63 companies the previous year), but only nine companies used long-term contracts for more than 50 percent of purchased gas on a peak day (compared to eleven companies the previous year). When asked to describe the distribution of gas supply purchases among suppliers, respondents cited independent marketers, producers and producing company affiliates more than any other class of supply aggregators. When asked if the company used asset management agreements for any portion of its gas supply purchases during the winter, 30 companies (41 percent) answered yes; however, 43 answered no. Supply Pricing Mechanisms and Hedging Issues Many factors play a role in the market pricing of natural gas and of transportation services, including weather, storage levels, end-use demand, financial markets and various operational issues. When asked to identify the tools most effective to managing supply and price risk, survey respondents largely cited physical storage and also mentioned fixed pricing (including advanced purchases at fixed prices), index pricing (both first of month and daily), and call and swing options. For long-term supplies (one year or more), 33 of the 42 companies used first-of-month (FOM) pricing for a portion of their supplies, including 21 companies that used FOM for at least 50 percent of long-term gas purchases. Sixteen companies utilized daily pricing, and 16 reported some form of fixed pricing. Mid-term purchases (more than one month, less than one year) were reported by 54 of 64 companies as most often tied to FOM indices for significant volumes of gas. In addition, daily mechanisms (28 companies), fixed-prices (27 companies) and NYMEX indices (15 companies--although for small volumes) were included in the mid-term pricing basket. Eighty-four percent of companies responding to the AGA survey (62 of 74) indicated that they used financial instruments to hedge at least a portion of their supply purchases for the winter heating season. This differs from two years prior when 92 percent (of a different sample of companies) indicated using financial hedging tools. In contrast, during the winter only 70 percent of survey companies used financial tools, while only 55 percent did so three years prior (during the winter). 4

5 For the WHS companies hedged as little as one percent and as much as 91 percent of winter heating season supply using financial instruments. The median supply volume hedged for the sample of companies was 36 percent. Options and fixed-price contracts were most often cited (by 33 and 30 companies, respectively) as hedging tools used for a portion of gas purchases. Other regularly used financial tools include swaps (20 companies) and futures (15 companies). The use of financial tools may be understated in this report inasmuch as some volumes delivered to LDCs from marketers and other suppliers are hedged by the third-party rather than the LDC or customer and may have been excluded from the LDC hedging calculation. Companies use a portfolio of timed hedges to balance their approach to strategic price planning. When asked about the strategic timing of their hedges, 50 of 62 companies (81 percent) indicated that they hedge 7-12 month forward for a portion of their supplies, while 49 of 62 companies employed a six-month-or-less timeframe. In addition, 28 companies used a 12-month-or-greater approach to hedge a portion of their supplies. Of these 62 companies that hedged supplies, 22 employed all three timing strategies. On the physical side in preparation for the WHS, 71 of 74 respondents (96 percent) reported using storage as a natural hedging tool. Thirty-nine of those companies hedged between 25 and 51 percent of winter heating season supplies using underground storage, compared with 33 companies last year. Another 21 companies employed this physical hedge for 1 to 25 percent of their supply portfolio. Only four of 73 survey respondents indicated that they used weather derivatives during the winter heating season. This compares with two of 51 companies two winter heating seasons prior. When asked about their own regulatory environment, 60 of the 61 companies that answered the question with an answer other than not applicable indicated that financial losses and gains tied to hedging were treated equally by the regulator. When asked about the focus of their regulator regarding gas purchases, 42 of the 64 respondents that knew the answer indicated that their regulator was interested equally in stable prices and the lowest price possible. Fourteen said that a lowest price was the only focus, while eight tagged stable prices as the concern. Only five of 72 companies indicated that regulators overseeing their services and activities in preparation for the winter heating season were less receptive to hedging than in the prior year. Three companies noted regulators as more receptive to hedging strategies, while 64 saw no change in regulatory receptivity to the practices and strategies in place. Gas Storage Production and market area storage are key tools for efficiently managing LDC gas supply and transportation portfolios. However, it should be noted that storage practices are no longer dictated solely by local utility requirements to serve winter peaking loads. Storage services now support natural gas parking, loaning, balancing, other commercial arbitrage opportunities at market hubs and city gates, and even supply resources during summer cooling periods. Sixty-eight of 71 companies (91 percent) indicated that weather-induced demand, among other factors, compelled them to utilize storage services. Respondents also cited no-notice requirements (58 companies), must turn contract provisions (45 companies), pipeline operational flow orders (28 companies), and arbitrage opportunities (21 companies) as 5

6 reasons to maintain storage services within their gas supply portfolio during the winter heating season. Must turn provisions may be in place for some storage contracts to maintain facility integrity through an optimal pattern of injection and withdrawal into and from a storage field. With that said, at the end of the prior winter heating season ( ), storage inventories finished much higher (nearly 2.5 Tcf of working gas remaining) than the prior five-year average and the prior year. In view of that, 62 percent of responding companies (38 of 61) said that mustturn provisions influenced their use of storage during the winter. Even though storage inventories finished the winter about 800 Bcf lower than the prior winter just cited, must turn provisions were still specified by 63 percent of the companies as influencing their use of underground storage. Sixty-three of the 70 companies that answered the question used first-of-month index pricing to purchase gas for injection into storage, and 70 percent (or 31) of those companies used FOM prices for percent of gas injected into storage. Forty-five companies indicated that they purchased a portion of their stored gas in the daily market; however, daily pricing tended to account for less than 25 percent of purchased storage volumes. Twenty-eight of 70 companies (40 percent) used fixed-price schedules for some portion of their storage purchases, compared to 48 percent the prior year. Eleven of 74 companies indicated that they were either constructing or studying the potential for adding underground storage during the next five years, while nine were considering adding market-area LNG or propane peak-shaving capacity to their gas supply assets of which one was in the process of building peak-shaving facilities. LDC Transportation and Capacity Issues Managing pipeline capacity efficiently is a challenge for many utilities and can involve the release of capacity to the secondary pipeline transportation market. From April 2012 to March 2013, 40 to 49 of the survey companies (varying with the month) released their unneeded pipeline capacity to the secondary market. Of those, 21 to 31 companies (depending on the month) released up to 25 percent of their pipeline capacity. During the spring-summer of 2012 (April through August), from nine to 13 surveyed companies per month released 26 to 50 percent of their capacity. Only 22 of 73 companies reported that operational flow orders (OFO) issued by pipeline companies, had an impact on their service territory during the winter heating season. The median number of these OFOs for these companies was 8.0. Only three companies reported consequential storage critical day issuances by system operators. 6

7 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Bcf per day III. Natural Gas Market Overview Why does a natural gas utility build a portfolio of natural gas supply tools to meet customer requirements during a given winter heating season? While the obvious reason is that companies want to deliver natural gas to customers reliably and at the lowest possible cost, another fundamental motivator is mitigating market uncertainty. Of course weather often introduces an element of the unknown for gas supply planners throughout the country. As a national trade association, AGA usually describes national natural gas markets, based on annual or monthly data. Since 1995 and up to 2009, U.S. natural gas consumption had been about Tcf annually, while U.S. natural gas production was about Tcf annually. By 2012 domestic dry natural gas production grew even more to 24 Tcf annually and consumption continued to rise. Even though these data indicate a level of stability and growth in the gas market, gas supply planners at local utilities face a very different picture --one that varies daily with fluctuating conditions that may turn extreme during winter heating season months. It is common knowledge that a balanced natural gas market is characterized by supply matching demand. Today s U.S. natural gas market balances consumption with domestic and international supplies at above 70 Bcf per day on average. However, on a daily basis during the course of a winter heating season natural gas consumption can fluctuate significantly. The graph in Figure 1 represents daily natural gas consumption from January through December 2012 and illustrates that winter heating season daily consumption does not necessarily correspond to annual or monthly averages. For example, from January 1 through March 31, 2012 daily natural gas consumption ranged from as little as 60 Bcf to over 100 Bcf. The graph also shows that consumption fell to below 60 Bcf per day for much of May but surged once again in July to meet natural gas-fired power generation requirements. FIGURE 1 U.S. Daily Natural Gas Consumption Source: Bentek Energy, LLC, Energy Market Fundamentals, December 31,

8 February March April May June July August September October November December Bcf per day Other physical flow and market fluctuations can be identified such as those seen in Figure 2, which shows net withdrawals from storage as a positive supply source and net injections as a demand requirement (below the zero line). Underground natural gas storage is in fact a valuable physical tool for managing sudden changes in weather-induced natural gas demand. Figure 2 Daily Storage Withdrawals (+) and Injections (-) January 1 December 31, Source: Bentek Energy, LLC, Energy Market Fundamentals, December 31, 2012 A look at the residential and small commercial sectors provides a sense of how extreme demand and consumption fluctuations can be on a day-by-day basis. Figure 3 (on the following page) graphs residential and commercial natural gas consumption data from January 1 through December 31, Here we see daily sector consumption as low as 16 Bcf for a warm winter day in March sharply contrasted with an over 50 Bcf consumption day in January and February. On a national basis, this represents more than a 100 percent load swing for natural gas utilities during the winter heating season. In most cases, changes in natural gas requirements are met with a package of supply tools including underground storage, peak-shaving facilities and others. For an individual utility this poses the ongoing challenge of meeting customer requirements each day of every winter and is the starting point for developing a portfolio of tools that are geared toward meeting this challenge. 8

9 FIGURE 3 Daily Residential/Commercial Natural Gas Consumption (Bcf) Source: Bentek Energy, LLC, Energy Market Fundamentals, December 31, 2012 IV. Weather Winter Heating Season According to data from the National Oceanographic and Atmospheric Administration (NOAA), the winter months were about 3.5 percent warmer than normal, following a winter season that had been 17.5 warmer than normal winter on a cumulative basis. The five-month winter season is unpredictable and can often demonstrate different patterns cold early, warm during its core or warm early and colder than normal in March. However, the 17.5 warmer than normal is practically off the chart, whereas the winter was a much more likely deviation from normal. Going back further both the and winter heating seasons were slightly warmer than normal based on heating degree day measures from October through March 0.2 percent and 1.5 percent, respectively. During the winter heating season, heating degree day totals varied from 14.4 warmer in December 2012 to 10.5 percent colder in March For the 22-week period of October 27, 2012 to March 30, 2013, eleven weeks presented colder than normal and 11 weeks were warmer than normal conditions, compared to NOAA s 30-year norm. Prior to the beginning of the winter season, NOAA had reported the first six months of 2012 as the warmest January through June on record in the lower-48 states. 9

10 On a regional basis for , cumulatively every region of the country ended up warmer than normal. Deviations from temperature norms for the various regions of the country varied from 1.2 percent warmer (West North Central) to 9.2 percent warmer (West South Central). TABLE 1 MONTHLY COMPARISON OF NATIONAL HEATING DEGREE DATA OCTOBER 2011 MARCH 2013 PERCENT CHANGE FROM NORMAL MONTH October 9.4% Warmer 3.2% Warmer November 13.1% Warmer 1.5% Warmer December 12.3% Warmer 14.4% Warmer January 18.0% Warmer 8.8% Warmer February 12.7% Warmer 2.0% Colder March 36.8% Warmer 10.5% Colder TOTAL 17.5% Warmer 3.5% Warmer Source: U.S. Department of Commerce, National Oceanic and Atmospheric Administration. V. Gas Supply Portfolios LDCs build and manage a portfolio of supply, storage and transportation services to meet expected peak-day, peak-month and seasonal gas delivery requirements. In today s business environment, gas portfolio managers continually attempt to strike a balance between their need to minimize gas-acquisition risks and their obligation to provide reliable service at the lowest possible cost. Given the reality of significant deviations from normal weather patterns (warm and cold) and regulatory scrutiny of costs to consumers, local gas utility exposure to hindsight analysis regarding gas supply practices is ever present. With that said, local gas utilities apply a standard or methodology for determining a design peak day temperature calculation, and this of course influences the construct of their gas supply portfolio. For the WHS survey, companies described their methodology for determining their design day calculation as follows: 19 employed a 1-in-30 year risk of occurrence, seven used a 1-in-20, three used a 1-in-15,, two used a 1-in-5, and two used a 1-in-10 year occurrence probability. Twelve companies utilized an alternative time period criteria, ranging from two years to 1-in-100 years. Twenty-eight companies indicated their use of other methodologies, such as historical peak adjusted for current and known changes, Monte Carlo statistical simulations, and weighted averages over specific time frames. Peak consumption predominantly occurred in January for survey respondents (70 of 74 companies). For these 74 companies, the aggregate peak-day sendout was 54.1 million Dekatherms during the WHS, making up 75 percent of the 71.8 million Dekatherms projected for peak-day requirements. Nearly 50 percent of respondents (36 of 74 companies) delivered between 60 and 80 percent of projected peak-day requirements. 10

11 Respondents were asked to depict their peak day and peak month delivered gas volumes by supply source. Table 2 and Figure 4 illustrate the diversity of gas supply sources available to LDCs. It should not be surprising that purchases moved by firm pipeline transportation provided much of the gas to consumers for the peak day and peak month during the WHS. Sixty-two of 74 companies indicated that firm pipeline supplies formed a part of their peak day gas supply portfolio, including 25 companies that showed 26 to 50 percent of their required peak-day volumes coming from firm supplies. Another 20 companies indicated that more than 50 percent of their peak-day supplies were moved via firm pipeline transportation. As shown in Table 2, also peak-month supplies were heavily weighted toward purchases via firm transportation. As with peak-day supplies, peak-month supplies tended to be supplemented with pipeline or other storage, city gate deliveries for transportation customers, city gate purchases, on-system underground storage, LNG or propane air, and local production. TABLE 2 SOURCES OF LDC PEAK GAS SUPPLIES WINTER HEATING SEASON (74 Companies) SUPPLY VOLUME PERCENTAGE RANGES CITY GATE PURCHASES CITY GATE SUPPLIES FOR TRANSPORTATION LNG PROPANE AIR LOCAL PRODUCTION ON-SYSTEM UNDERGROUND STORAGE PIPELINE OR OTHER STORAGE PURCHASES MOVED VIA FIRM PIPELINE TRANSPORTATION PURCHASES MOVED VIA INTERRUPTIBLE TRANSPORTATION OTHER PEAK DAY 1 25% PEAK MONTH 1 25% Table 2 and Figure 4 also demonstrate that companies tend to diversify their supply strategy in increments that often amount to less than 50 percent of their total supply package. Besides firm pipeline transportation, other categories of gas supply were also important for peak-day deliveries by the sample companies: 57 of 74 companies indicated that up to 50 percent of peak-day supplies originated from pipeline or other storage, while 51 companies indicated that up to 50 percent of their peak-day supplies were city gate supplies for transportation customers. Twenty-one made city gate purchases, 20 used on-system storage, 16 used LNG or propane air as a supply source, and 11 11

12 NUMBER OF COMPANIES utilized local production. This year, as many as four respondents used interruptible transportation for their peak deliveries, whether on a peak day or within a peak month. Last year, no company reported interruptible transportation as a peak day or peak month supply source. The other category, reported by nine companies, includes pipeline operational balancing agreement receipts, line pack and draft, off-system transport, and interstate supplies. FIGURE 4 40 Sources of Peak-Day Gas Supplies Winter Heating Season (74 LDCs) % 26-50% 51-75% % SUPPLY VOLUME PERCENTAGE RANGES Citygate Purchases Citygate for Transp. Customers LNG Propane-Air Local Production On-system Underground Storage Pipeline Storage Firm Transportation Interruptible Transportation Other Supply diversity is not limited to the gas source. Local gas utilities also employ a diverse set of contractual arrangements to procure their gas supplies, including long-term, mid-term, monthly and daily agreements. A mix of contracts allows the LDC to balance between competing needs, such as the obligation to serve its customers as the supplier of last resort and the need to maximize efficiency while minimizing costs. In many cases, longer-term contracts contribute to baseload obligations, while shorter-term contracts allow companies to respond to market changes. However, the recent waning of market volatility, particularly as it applies to natural gas acquisition prices, is resulting in a reexamination by consumers and regulators of supply acquisition contracting, with less emphasis on absolute least cost and more stress on price stability. Some argue that longer-term contracting may be useful to underpin new supply sources in the future. Generally the data show a balance among contract lengths of peak-day and peakmonth supply volumes, particularly for volumes up to 50 percent of requirements (see Table 3). However the use of mid-term deals (defined as greater than one month and less than one year) is becoming more prominent for gas volumes above 50 percent of gas requirements. Table 3 includes 12

13 contract terms for winter heating season supplies, and it shows a slight increase in monthly agreements for the five-month winter period compared to the peak-day and peak-month. For WHS peak-day supplies, long-term agreements (defined as one year or longer) were used by 39 of 72 companies (compared to 36 of 61 last year). Of those, nine companies used long-term contracts for more than 50 percent of their peak-day supplies. In comparison long-term deals were made for more than 50 percent of peak-day gas purchases by fourteen of the WHS survey companies. However, eleven of last year s survey companies acquired more than 50 percent of their peak-day supplies via long-term contracts. TABLE 3 CONTRACT TERMS FOR GAS PURCHASES WINTER HEATING SEASON SUPPLY VOLUME PERCENTAGE RANGES DAILY LONG-TERM (> 1 YEAR) MID-TERM ( 1 MONTH > 1 YR) MONTHLY OTHER PEAK DAY (72 COMPANIES) 1 25% PEAK MONTH (72 COMPANIES) 1 25% WINTER SEASON (71 COMPANIES) 1 25%

14 Mid-term deals for peak-day purchases were made by 62 companies during the WHS more companies than those using daily (44 companies), long-term (39 companies), or monthly arrangements (36 companies). A similar pattern emerges for peak month and winter season purchases, with shorter-term deals (daily and monthly) more represented over the winter season, particularly for volumes less than 25 percent of gas purchase requirements. The other category includes long-term pre-pay deals, storage withdrawals and other arrangements. As to supply providers, as shown in Table 4, when asked to describe the distribution of peakday gas purchases among suppliers, 57 LDCs identified independent marketers. The balance of supplies acquired by LDCs were distributed among producers (46 companies), producing company affiliates (33 companies), LDC energy marketing affiliates (19 companies), pipeline energy marketing affiliates (14 companies), pipeline companies (three companies), and LDC-owned production (two companies). The other category includes financial marketing affiliates, other utilities, asset managers, storage operators, power generators, electric utilities, end users, and other supply aggregators. TABLE 4 DISTRIBUTION OF PEAK GAS PURCHASES AMONG SUPPLY PROVIDERS WINTER HEATING SEASON (71 COMPANIES) SUPPLY VOLUME PERCENTAGE RANGES INDEPENDENT MARKETER LDC ENERGY MARKETING AFFILIATE LDC OWNED PRODUCTION PIPELINE PIPELINE ENERGY MARKETING AFFILIATE PRODUCER PRODUCING COMPANY AFFILIATE OTHER PEAK DAY 1 25% PEAK MONTH 1 25% When asked whether their company used asset management agreements for any portion of its gas supply purchases during the winter heating season, 30 of 73 companies (41 percent) said yes the same percentage as the prior winter heating season. Of the 29 companies that answered more specific questions about third-party asset management, 14 used asset management for 25 percent or less of their winter heating season supplies, while 10 of the companies utilized asset managers for more than 50 percent of winter heating season supplies (see Table 5). 14

15 TABLE 5 PORTIONS OF WINTER HEATING SEASON ACQUISITIONS VIA ASSET MANAGEMENT AGREEMENTS (27 COMPANIES) SUPPLY VOLUME PERCENTAGE RANGES NUMBER OF COMPANIES 1 25% VI. Supply Pricing Mechanisms and Hedging Pricing Mechanisms Many factors play a role in the market pricing of the gas commodity and of transportation services, including weather, storage levels, end-use demand, pipeline capacity, operational issues, and financial markets. The market fundamentals that impact price have also expanded to include interest rates, other investment opportunities, the price of other commodities and even currency exchange rates. Such broad market influences impact LDCs and other gas suppliers, making planning increasingly difficult for all stakeholders. In order to deal with the inherent uncertainty of the market even considering the relative stability of natural gas markets in recent years supply planners use a portfolio approach to pricing gas supplies mirroring their approach to supply sources, providers and transportation options. Along with pricing mechanisms and contract terms noted below, the notion of adding fixedprice longer-term supply contracts to supply portfolios has resurfaced as a value-added tool for managing price stability in today s market. Future key gas supply projects, such as those aimed at coordinating natural gas and power generation loads, may require longer-term demand pull contract arrangements to be successful. When asked whether they would consider including fixed-price supply deals in their 1-3 year term supply contracts, at a price in a $5-6 per MMBtu range, if regulators would approve such deals, 22 of 66 companies (33 percent) said yes, and 31 said maybe. Two years prior, half the survey companies (25 of 50) answered yes to the same question. This year s responses possibly reflect a general feeling that Henry Hub prices will likely remain stable overall but at even lower levels than previously anticipated. Of the 22 companies that answered yes to the hypothetical question this year, 16 chose a percentage range of 1 20 percent of their total supply for longer-term fixed-price arrangements. Two opted for percent, two elected percent, and one chose percent of supply volumes. Only one company indicated that it would build over 50 percent of its total supply portfolio on long-term, fixed-price deals (in this case, percent). With respect to preferred contract durations for such deals, 11 of the 22 companies found 1-2 year terms as optimal, eight favored terms longer than two years, and three preferred less than one year. Seven of the 31 companies that answered maybe regarding longer-term fixed price arrangements said they would consider percent of their supply purchases for such deals, six would opt for 1-10 percent, eight would look at percent of supply volumes, and six would consider more than 30 percent. With regard to contract durations, 13 of these companies view 2 years or more as optimum, four favor 1-2 years and three prefer less than one year. 15

16 When examining the natural gas purchasing practices of companies during the past several winter heating seasons, it is clear that first-of-month (FOM) index pricing dominates the market for the largest portion of supply agreements, whether short, long or mid-term. Table 6 provides a closer look at the balance of pricing mechanisms among survey respondents during the winter heating season. TABLE 6 GAS SUPPLY PRICING MECHANISMS WINTER HEATING SEASON COMPANIES OVERALL SUPPLY VOLUME PERCENTAGE RANGES AVERAGE LAST 3 DAYS DAILY (SPOT OR INDEX) FIRST-OF- MONTH INDEX FIXED NYMEX WEEKLY OTHER LONG TERM (GREATER THAN 1 YEAR 42 COMPANIES) 1 25% MID TERM (1 MONTH > 1 YEAR 64 COMPANIES) 1 25% SHORT TERM (1 MONTH OR LESS 58 COMPANIES) 1 25%

17 NUMBER OF COMPANIES As shown in Table 6 and Figure 5, 33 of the 42 companies with long-term supplies (one year or more) used first-of-month pricing for a portion of these supplies, including 21 companies that used FOM for at least 50 percent of purchases. Sixteen companies used daily pricing mechanisms for long-term supplies, and most of them used this pricing for less than 50 percent of their supply volumes. Also 16 companies utilized some form of fixed pricing (compared to 12 the prior year). By way of comparison, for the WHS 15 of 47 companies used fixed pricing, while ten years ago, only 10 of 40 companies cited fixed deals. FIGURE 5 16 LDC Long-Term Gas Supply Pricing Mechanisms Winter Heating Season (42 LDCs) % 26-50% 51-75% % SUPPLY VOLUME PERCENTAGE RANGES Average Last 3 Days Daily First-of-Month Index Fixed NYMEX Weekly Other Figures 5, 7 and 8 show the pricing mechanisms employed by this year s survey participants, and Figures 5 and 6 together present a comparison of long-term pricing arrangements for the past two winter heating seasons. The graphs clearly show that for larger volumes of gas purchased under long-term arrangements, first-of-month indices continued to be the predominant pricing mechanism during just as they were for the winter. This is not surprising, since the first-of-month index is not only a measure of market movement but also often serves as baseline from which hedging strategies can be measured. Fixed pricing also played a somewhat more prevalent role for larger long-term volumes relative to other mechanisms. The relative prominence of these two pricing mechanisms may be explained by the relative price stability that appears to have developed in the natural gas market recently, given an overall strong natural gas supply position based on six consecutive years of growth in domestic production. Weekly and average three-day pricing played no role in long-term gas purchases during the WHS. 17

18 NUMBER OF COMPANIES FIGURE 6 LDC Long-Term Gas Supply Pricing Mechanisms Winter Heating Season (37 LDCs) % 26-50% 51-75% % SUPPLY VOLUME PERCENTAGE RANGES Average Last 3 Days Daily First-of-Month Index Fixed NYMEX Weekly Other According to the 64 companies that reported mid-term supplies (of more than one month and less than one year) during the WHS, much of these natural gas purchases were tied to FOM indices (54 companies, including 35 that used FOM pricing for more than 50 percent of their supply). However, as Table 6 and Figure 7 indicate, daily, NYMEX and fixed pricing mechanisms were used to a significant extent for smaller-volume mid-term purchases. Twenty-seven companies reported using fixed pricing mechanisms for mid-term purchases, compared with 16 for long-term purchases. Also 28 companies used daily prices for mid-term purchases, compared with 16 for longterm purchases. As would be expected, more companies (44 of 58) used daily pricing for short-term purchases (one month or less) during the WHS, than for mid-term or long-term purchases; however, these short-term purchases were also heavily dependent on first-of-month indices (34 companies) and tied to fixed prices and NYMEX indices (see Table 6 and Figure 8). It should be noted that LDCs build gas supply portfolios and pricing strategies based on prior as well as anticipated experiences. Even state regulator-approved pricing mechanisms may appear favorable one year while less so the next. Flexibility and constructive policy reviews, rather than secondguessing, can have a positive effect on the delivery of natural gas and services to customers at the lowest possible cost. 18

19 NUMBER OF COMPANIES NUMBER OF COMPANIES FIGURE 7 25 LDC Mid-Term Gas Supply Pricing Mechanisms Winter Heating Season (64 LDCs) % 26-50% 51-75% % SUPPLY VOLUME PERCENTAGE RANGES Average Last 3 Days Daily First-of-the Month Index Fixed NYMEX Weekly Other FIGURE 8 18 LDC Short-Term Gas Supply Pricing Mechanisms Winter Heating Season (58 LDCs) % 26-50% 51-75% % SUPPLY VOLUME PERCENTAGE RANGES Average Last 3 Days Daily First-of-the Month Index Fixed NYMEX Weekly Other 19

20 Hedging Mechanisms Market developments during and since the 1990s have expanded the options for acquiring gas supply, trading transportation capacity, and using financial instruments. Today industry players use futures contracts and other tools to offset the risk of commodity price movements. These financial instruments, which include fixed-price gas purchase contracts, futures, swaps and options, allow gas supply portfolio managers to hedge or lock in a portion of the commodity cost component of gas supplies. This is accomplished well when the required level of risk and the rewards or benefits of managing such risk are properly balanced by the company, consumers and regulatory bodies. Eighty-four percent of responding companies (62 of 74) said they used financial instruments to hedge a portion of their winter heating season gas supply purchases. This percentage is slightly higher than last year, when 81 percent of companies indicated using financial hedges, but it is lower than for the past three years, where 92 percent of companies reported using financial tools in , 90 percent in , and 89 percent in Still this percentage is significantly larger than in (70 percent of respondents) and in , where only 55 percent of respondents reported using financial tools to hedge gas supply costs. It is important to note that the company makeup and size of the survey sample differ from year to year. For the winter, 49 of 60 responding companies hedged up to 50 percent of their gas supply purchases (compared with 34 of 47 companies for the winter). Respondents used one or more of the following instruments to hedge a portion of their WHS gas supply purchases: options (33 companies), fixed price contracts (30 companies), swaps (20 companies), and futures (15 companies). The use of financial instruments may be understated in this report inasmuch as some of the volumes delivered to LDCs from marketers and other suppliers are hedged by a third-party rather than the LDC and may have been excluded from the LDC s data. Only four companies reported using weather derivatives during the winter heating season. This compares with two of 51 companies in , five of 76 companies in , and seven of 54 in the survey. When asked about the timing of hedging strategies, 49 of 62 of the companies with hedging programs (79 percent) indicated that they applied a six-month or less strategy for a portion of their hedges for the winter heating season. Fifty companies used a 7-12 month strategy, and 28 companies employed a greater than 12-month strategy. Of course, a single company may use one or all strategies simultaneously. In fact, 24 of the respondents did just, compared with 19 the prior year. On the physical side, companies view gas supplies delivered to storage during the summer refill season as a price hedge against potential winter price run-ups. In preparation for the winter heating season, 71 of the 74 reporting companies (96 percent) used storage as a physical hedge. Sixty companies reported using storage for up to 50 percent of winter heating season supplies, compared with 52 and 46 companies for the and winter heating seasons, respectively. In some jurisdictions there are no formal standing hedging plans. In others, LDCs may be required to have their hedging plans for future gas supplies in place by predetermined dates. Variations on these themes are many and are geared to be compatible with the interplay among local distribution company, regulator, and local market conditions. Of the 74 reporting companies, 22 noted that their regulator required a hedging plan to be filed for approval. In addition, 24 companies indicated that state regulators place restrictions on hedging parameters, such as the choice of financial tools, date ranges, and/or the quantities hedged. Of these companies, seven indicated that their regulator requires both a plan and restrictions on 20

21 hedging. Thirty-five of the respondents noted that no plans or restrictions were required for their programs. When asked about their regulatory environment, the majority of respondents (64 of 72) reported no change in their regulator s receptivity to financial hedging during the winter heating season compared to the prior year, and three reported increased receptivity on the part of their regulator or public utility commission (PUC). Five companies indicated that their PUC was less receptive this past winter heating season. Sixty of 61 responding companies reported that their regulator treated the financial losses and the gains related to hedging equally. This near-100 percent response compares with 88 percent (or 45 of 51 companies) two years ago, 81 percent the year prior, and 78 percent the year before that. Additionally, 60 companies answered yes when asked if costs associated with their financial hedging programs were fully recoverable, while two respondents reported that it is some of the time but not guaranteed. When asked about the focus of their regulator with respect to natural gas purchases, fourteen respondents indicated that their regulator was primarily interested in the lowest possible price, eight said that the focus was on stable prices, and 42 companies said their regulator was equally concerned with both low and stable prices. Among LDCs, motivations vary surrounding hedging programs. When asked how customers benefited from their financial hedging compared with no hedging, 41 of 62 companies (66 percent) noted the reduced volatility in prices as a major benefit to customers, two cited reduced gas costs as the main advantage to customers, and fifteen observed both effects for their customers. VII. Gas Storage As noted earlier, local distribution companies are concerned with managing gas supply and transportation portfolios efficiently and cost effectively. Production area storage and market area storage help LDCs meet these goals. The use of storage facilities helps LDCs to both meet shortterm swing opportunities and satisfy peaking needs. Table 7 shows storage levels as estimated by the Energy Information Administration for January-April 2012 compared to the same period in For the nation as a whole, working gas inventories in both years were not strained. By the time net injections began in earnest in mid-april 2012 working gas levels were at record highs for a season s end about 2.5 Tcf. This could be attributed to a winter that was practically a non-event at 17.5 percent warmer than normal. In contrast, at season s end in the spring of 2013, winter conditions were slightly warmer than normal and accordingly underground storage was more fully utilized, having drawn down to about 1.7 Tcf. The strong March 2012 ending storage inventories resulted in 7 Bcf per day of net injections during that summer instead of the starting point of Bcf per day of injections with most of the balance going to power generation during a very warm summer and almost no influence on market acquisition prices. Put mildly, 2012 was an extraordinary year for natural gas markets. 21