THESlS This is to certify that the thesis entitled QUALITY INDEXES AND GREEN LUMBER YIELDS FOR GRADED MICHIGAN RED OAK SAWLOGS presented by DAVID THOMAS BOZAAN has been accepted towards fulfillment of the requirements for $1.5. degree in Forestry \ Maj ofessor Date 7,104) 8:, ,9 Q3 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES ”- REIURNING MATERIALS; Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. T____._._...- QUALITY INDEXES AND GREEN LUMBER YIELDS FOR GRADED MICHIGAN RED OAK SAWLOGS By David Thomas Bozaan A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Forestry 1983 ABSTRACT QUALITY INDEXES AND GREEN LUMBER YIELDS FOR GRADED MICHIGAN RED OAK SAWLOGS BY David Thomas Bozaan The yield of grade lumber from 404 graded red oak logs was measured, and the relative price for each lumber grade was calculated to develop quality indexes (QI) with regard to log diameter and log grade. Actual yields were tabulated then regressed on log diameter to produce curved yields. Timbers and crossties made up 25 to 80 percent of the lumber recovery volume depending on log grade. Grade 1 logs produced a mix of grade lumber with a relative value, that is, an average 01 of 1.19, or 119 percent that of No.1 Common red oak lumber. Average 015 for grade 2, grade 3, and grade 4 red oak logs were .89, .61, and .46, respectively. The 015 can be used to determine the value of sawn lumber produced from graded logs (relative to an equal volume of No.1 Common lumber), but may require some individual adjustment to improve accuracy. 950)»)57 §~:§‘IJL_'L.-"II Without wood a fire goes out .... Proverbs 26:20 To My Family, Who fueled my efforts ii ACKNOWLEDGEMENTS Many thanks to my advisor, Dr. Henry Huber, for his helpful guidance and for providing the support needed to undertake this study. I'm indebted also to committee members Dr. Alan Sliker and Dr. John Hart for their willingness to share ideas and patient criticisms throughout my degree program. To Dr. Carl Ramm, for help in statistical matters. Special thanks to Jack Zollner of the Michigan Department of Natural Resources for his critical involvement in field planning and data collection, and for sharing his vast wealth of experience in the Forestry and wood-using professions. The author appreciates help given by the U.S. Forest Service at Bromall, Pennsylvania in conducting the SOLVE computer analysis. To Tim Youngstrom and Don Michael, thanks are in order for their detailed help in the computer programming vital to this study. John Heckman and Bob Freyman are also acknowledged for their helpful suggestions. My sincere thanks and heartfelt appreciation to Maria Pate whom, as friend and typist, helped humor many late hours of manuscript preparation. Lastly, my parents and family deserve my deepest gratitude for their loving support and encouragement throughout my Masters education. iii LIST OF TABLES. . . LIST OF FIGURES . . INTRODUCTION. . . . LITERATURE REVIEW . TABLE Development of Log Grades Quality Index Development Lumber Recovery from Graded Logs. Log Characteristics PROCEDURE . . . . . Mill Selection. . Data Collection at the Sawmill. SOLVE Analysis. . Grade Yield Tabulation. Price Relative (p. r. ) Calculation Quality Index (01) Calculation. RESULTS AND DISCUSSION. Actual Grade Yields Curved Yields and Regression Equations Log Characteristics Price Relatives . Quality Indexes . OF CONTENTS Comparison of Estimated Value of Tallied Lumber against 01- estimated Log- product Value. . . . . . Overrun Factors: QI Adjusted for Other Log Rules . CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . LITERATURE CITED. . APPENDICES A. NUMBER OF LOGS, NET DOYLE LOG SCALE, MBF LUMBER TALLY, AND PERCENT OVERRUN FOR EACH SAWMILL, BY LOG GRADE . . iv Page vi viii ._a._a W—SUWLA’ w 15 15 15 16 17 18 19 21 21 26 A1 43 50 52 5M 57 61 APPENDICES (continued) B. QUARTERLY ”/4 RED OAK LUMBER PRICES AND FAS/SEL PREMIUMS’ 1978-83 0 o o a o o o o o o o o 62 QUARTERLY PRICE RELATIVES, 1978-83. . . . . . . . 63 FOREST SERVICE STANDARD GRADES FOR HARDWOOD FACTORY LUMBER LOGS O O O C O O O O C C O O O O O 6” Table 1. 2. 10. 11. 12. 13. 14. Actual green sawlogs, in Actual green sawlogs, in Actual green sawlogs, in Actual green LIST OF TABLES lumber yields percent . . . lumber yields percent . . . lumber yields percent . . . lumber yields for red —I oak grade oak grade 2 oak grade 3 oak subfactory class sawlogs, grade A, in percent. . . . . . . . Comparison of actual red oak lumber yields for study logs from four sawmills with expected FPL-63 f0!” the same 1088. o o o o o 0 yields from Curved green sawlogs, in Curved green sawlogs, in Curved green sawlogs, in Curved green class sawlogs, grade 4, lumber yield for red oak grade 1 percent . . . lumber yield for red oak grade 2 percent . . . lumber yield for red oak grade 3 percent . . . lumber yield for red oak subfactory in percent. . . . . . . . Regression equations, mean lumber yields, standard error of estimates, and adjusted coefficients of determination (R lumber grade. . . . . . . Price Relatives: ) for red oak logs, by log and Comparison of 1978-83 study results with 1970-74 and 196H-68 findings . . . . Quality indexes for red oak by log grade and diameter class. . . . . . Quality indexes for red oak by log grade: A comparison of six studies from 1952-1983. . . . Comparison of estimated value of tallied lumber from four study sawmills with QI-estimated log product value, by log grade . vi Page 22 23 24 25 27 3O 31 32 32 33 “A 45 M7 51 Table Page 15. Red oak quality indexes adjusted for International 1/8 and Scribner Decimal C log rules. . . . . . . 53 16. Overrun factors by log rule and log grade, in percent 0 O O O O O O O O O 0 O O O O O O O O O O 53 vii LIST OF FIGURES Figure Page 1. Cumulative distribution of lumber grades for red oak sawlogs, by log grades. . . . . . . . . . 28 2. Percent distribution of red oak sawlogs, net Doyle scale, and lumber tally, by log grade (”on log sample). 0 I O O O O O O O O O O O O O O 37 3. Distribution of log diameters for graded red oak sawlogs, in percent (40“ log sample). . . . . . . 38 u. Length distribution for graded red oak sawlogs, in percent (40” log sample) . . . . . . . . . . . MO 5. Curved lumber yields for red oak grade 1 and grade 2 sawlogs, in percent . . . . . . . . . . . 48 6. Curved lumber yields for red oak grade 3 and grade A sawlogs, in percent . . . . . . . . . . . “9 viii INTRODUCTION The concept of a hardwood log quality index (QI) was developed by Allyn M. Herrick in 1946 as a measure of log quality and value. The index estimates a log's value before it is sawn, relative to an equal volume of No.1 Common lumber. To calculate the product value of a log using QI, one must know: (1) the quality index for the desired species; (2) the net log volume; and (3) the current market price of 4/4 (one-inch) No.1 Common lumber. Multiplying the three variables results in an estimate of the value of the products that will be sawn from the log, assuming a mill will saw standard lumber. The quality index is directly related to the lumber recovery from graded hardwood logs. Grade yield information can be found in U.S. Forest Service publications FPL-63 and NE-468 (14,36), or can be developed for a particular region or species. In Michigan, red oak is of particular interest to lumbermen and researchers alike because of its desirable wood characteristics and economic importance. During 1978, 93.5 million board feet (MMBF), or 16.6 percent of Michigan's sawlog production, was made up by red oak logs (5). Only aspen and hard maple sawlog production surpassed that of red oak, composing 22.6 and 18.5 percent, respectively, of l Michigan's total output in 1978 (5). Ninety-seven percent of the red oak sawlogs were produced from Michigan's Lower Peninsula. In a previous study by Huber, Bozaan, and Draper (18), 14 sawmills in central Michigan were studied to determine their internal cost structure. Seven of these mills sawed primarily graded hardwood lumber, of which four sawed a large proportion of red oak sawlogs. For this study, the author undertook to examine closely grade lumber yields from mill-run mixes of red oak sawlogs from these four sawmills. Using this grade yield information, quality indexes were developed to estimate potential log-product values. The- results are reported and discussed in the following text. LITERATURE REVIEW Dexelcpment Of L22 Grades Early in this century, the ability to accurately measure hardwood log quality was seen as an important means of predicting grade lumber yields and value. Judging a log's quality and expected output was a matter left to lumbermen and sawyers with years of experience. This concept of log "grading" changed in the mid 19305 as the U.S. Forest Service needed to assess log quality fOr the purpose of surveying standing volumes and potential yields of Lake States hardwoods. H.G. White (38) of the U.S. Forest Service Lake States Forest Experiment Station presented a paper at the 1937 meeting of the Michigan Academy of Science, Arts, and Letters advocating the need for an accurate and selective measure of timber quality. He indicated that softwood log quality had been studied for years in the West, but that little work had been done on hardwoods in the East which were more variable in quality than softwoods and had defects that were more difficult to appraise. White emphasized that the wide range of values between grades of hardwood lumber made grading of log quality essential. White proposed log specifications in the form of three U.S. Forest Service log grades which had been used for the 3 Forest Survey of the Lake States. The grades were based on two types of observations: (1) the size and appearance of the log; and (2) the output of lumber from the log, especially No.1 Common and Better lumber. A list of species surveyed and their percentages of grade 1, 2, and 3 logs were presented. In 1949, hardwood log grade specifications were published by the Forest Service's Forest Products Laboratory in a paper titled "Hardwood log grades for standard lumber -- proposals and results" (2). This system was directly related to National Hardwood Lumber Association rules for standard grades of lumber (34). In 1952,, the proposed grades were adopted as the official hardwood log grades for the U.S. Forest Service (14). W.S. Bromley (6) reviewed White's work discussing the development of log grades by the U.S. Forest Service for determining relative quality of logs and trees. He also reviewed a Forest Service logging and milling study conducted at Phelps, Wisconsin, which showed marked differences in values between respective log grades. Bromley's study evaluated whether the grade of logs removed from a partial cutting on Ford Motor Company lands in Alberta, Michigan, could be used to estimate or appraise the value of lumber sawn from them. The logs in his study were graded and scaled, then yield percentages from the Phelps, Wisconsin, study were used to calculate estimated lumber value. These estimated values were then compared with actual values of the 5 sawn, graded, and tallied lumber at the mill. He found that the log grades gave good estimates of value for two of the three species studied, but suggested that the procedure was too tedious to be flexible in the event of changes in lumber price. A simpler method for estimating value was needed. QualitxlndexDemment In 1946, Allyn M. Herrick (16) studied grade yields and overrun from Indiana hardwood sawlogs in an attempt to estimate product value. He saw the need for a single expression that would denote average quality of a tree or log as measured by the grades of lumber it would yield. Herrick developed the hardwood log quality index (QI) to fit this need. The first quality index estimated a log or tree's value relative to an equal volume of First and Seconds (FAS) lumber. Two types of information were needed to calculate 01: (1) the percentage yield of each lumber grade sawn from a log; and (2) the ratio of each lumber grade price to that of an FAS reference price. These price ratios, later called price relatives, were found by dividing market price for each grade of 4/4 lumber ($/MBF) by the price of FAS lumber ($/MBF). When yields and price ratio (p.r.) for each respective lumber grade are multiplied and summed, a quality index for a particular species results. Herrick used only the four basic lumber grades: 01 = (p.r. FAS)(%FAS) + (p.r. No. IC)(1No. IC) + (p.r. No. 2C)(%No. 20) + (p.r. No. 3C)(%No. 3C) His log quality index became a new means of expressing average lumber grade recoveries and value. As a result of Herrick's work, several papers on quality index research were presented at the Symposium for a Standard Hardwood Quality Index at Purdue University in 1952 (7,8,21,23,37). Wallace (37) found that QIs did not increase regularly with increasing log diameter for uncurved yield data. (A line of best fit calculated by this author for the quality index values he published showed only a slight increase in 01 with diameter.) He surmised that the variation in QI must be directly related to the variation in lumber grade recovery for logs of different sizes. Campbell (7) pointed out that higher-than-estimated quality indexes are necessary when sawing better tree grades with a band mill. Band mills often saw 6/4 lumber. Cummins (8) presented work demonstrating the historical and statistical stability of price relatives. He found nonsignificant variation of price relatives among the three Horgwooo Market Report regions (Southern, Appalachian, and Northern) over a number of years, although the Southern market p.r.s were consistently lower than the other two. Using No.1 Common as the reference grade, Cummins found that the value of lumber estimated using p.r.s from a five-year base period was within 8 percent of actual value for two- thirds of the examples studied. Kramer (21) noted that the quality index formula is flexible. It may be expanded or reduced to compensate for the number of lumber grades being sawn in any given locality. 7 Lockhard (23) found that 015 are somewhat more uniform in the lower grade logs. Based on these findings, the Symposium for a Standard Hardwood Quality Index adopted the following standards for calculating quality indexes: 1) No. 1 Common would be the standard reference grade. 2) 4/4 (l-inch) lumber would be the standard reference thickness. 3) p.r.s would be calculated from average lumber prices over a 5-year base period. Beazley and Herrick (3) published work on lumber price relatives and their application in the hardwood quality index. They determined that quarterly rather than weekly lumber price quotes, were sufficient for calculating p.r.s for a given five-year base period and that any five-year base period would provide adequate price information so long as it included no major economic upheaval such as the Great Depression of the 1930s or World War II. They also demonstrated the first use of No. 1 Common as a standardized reference grade for calculating p.r.s (3). Herrick (17) evaluated the application of quality index in hardwood sawtimber management. He reviewed studies by Ellersten and Lane (9) and Beazley and Herrick (3) showing that lumber price relatives for different grades are quite stable regardless of changing grade lumber prices and concluded that an index of quality based on p.r.s should be superior to actual prices for estimating lumber values from logs and trees. Herrick proposed that, although species differences existed, pooled p.r.s averaged for nine hardwood species were useful for calculating a quality index for any hardwood species in the Central States in 1956. Additionally, he found log grade and log size to be more important determinants of QI than was species. Bentley and Streeby (4) outlined procedures for projecting quality indexes ten years into the future based on previous 10-year price relative and market price trends. Prices from the base period 1957-66 were used to project 015 for the 10-year period 1967-76. Quality indexes were published for three species and two log rules, International 1/4 and Scribner Decimal C. They raised an important issue_ not previously investigated: that 015 averaged over diameter must be adjusted for log rule because of the different overrun factor involved with each rule. Streeby and Bentley also developed a computer program named QUINDEX for estimating and projecting quality indexes for different log grades (33). McCauley and Mendel (25) compared actual sawlog values at seven sawmills in southeastern Ohio and eastern Kentucky with quality index values developed for the study. Comparisons showed that for all species in log grades 1 and 2, commercial sawlog product values on the average were 4.9 percent less and 7.3 percent greater, respectively, than those predicted by the quality index for the same logs. Red oak commercial sawlog product values for the same grades were, respectively, 3.4 percent less and 22.7 percent greater than those predicted by the quality index. 9 The authors made two significant generalizations for grade 1 and 2 logs: (1) sawmill operators studied tended to saw logs into products valued within plus or minus 10 percent of that predicted by the 01, and (2) a quality index based on 4/4 lumber prices was a reliable predictor of grade 1 and 2 log product values in 1962. For logs below grade 2 (grade 3, construction, and local use), commercial log product values were about 50 percent greater than predicted by OI, and an according adjustment was suggested. Trimble and Mendel (35) used the quality index concept to calculate log values, in terms of amount and value of lumber they contained, for determining financial maturity of trees of three oak species. The log values were used to calculate present value (V0) of individual trees, which were in turn used to find prospective tree value over a ten-year period. Mendel and Smith (27) published quality index tables for eight eastern hardwood species by log grade and diameter class for two market reports, The Harowooo Market Report and The Qommeroial Bulletin. The 015 were calculated from price relatives averaged over the base period 1964-68. Air-dry grade lumber yield information was taken from the Forest Products Laboratory publication FPL-63 (36) and the actual yields were hand curved by the authors to reduce their erratic nature. Huyler (19) used the log quality indexes published by McCauley and Mendel (25) to compare lumber value of grade 3 live-sawn red oak logs with that of grade-sawn logs. The 10 quality index for each log diameter and grade was multiplied by the current market price of 4/4 No. 1 Common lumber and by the log's volume to find lumber value: Lumber Value = OI of log x price of No. 10 lumber x volume of log In contrast to Mendel and Smiths (27) hand curving procedure, actual air-dry yields from FPL-63 were used for the QI calculations. Air-dry lumber volume was estimated by deducting 5 percent from the green volume of red oak lumber. Comparing price relatives from two base periods, Mendel and Peirsol (28) found a need to update price relatives regularly to maintain accurate QIs. Their study showed that p.r.s change as a result of fluctuating supply and demand for the different grades of lumber and subsequent changes in lumber price. Price relatives and 015 were calculated for the base period 1970-74 for eight eastern hardwood species to update Mendel and Smith's (27) earlier work, which used 1964- 68 as a base period. The later study found that 013 had changed significantly since the 1964-68 base period for four of the eight species studied, including red oak. The average quality index value for red oak had dropped -.07, -.04, and -.04, respectively, for the three U.S. Forest Service factory hardwood log grades 1, 2, and 3. Thus, the use of 1964-68 013 would have overestimated the value of lumber sawn from grade 1, 2, and 3 red oak logs for the 1970-74 base period. Johnson and Meteer (20) published lumber grade yields and quality indexes for five northern hardwood species from ll logs graded under the Northern Hardwood and Pine Manufacturers Association (NHPMA) Log Grading Rules (29). Grade yields were reported as rough-green lumber with no allowance for shrinkage. The authors calculated price relatives using price quotations from a single weekly HELQHQQQ Ménkfii BQEQLL (February 24, 1979) instead of the 5- year base period recommended by the Symposium on a Standard Hardwood Quality Index. Wilson et al (39) presented quality indexes for five northern hardwood species in conjunction with uncurved yield data. Their tables indicated that, although red oak QIs did increase with diameter, the increase was not regular. Lumber £3.29an from Graded Loss In 1949, the U.S. Forest Products Laboratory published information on the relationship between log characteristics and end-product yield (2). Yields of standard lumber for proposed U.S. Forest Service log grades were presented. In 1966, this work was updated with extensive green lumber yield information for 14 hardwood species (36). Approximately 11,000 logs were sawn at 28 sawmills in northern, central, and southern hardwood regions (36). Schroeder and Hanks (31) compiled lumber grade yield information for 556 factory grade red oak sawlogs sawn at four sawmills in West Virginia and Virginia. Tables showing actual and curved grade yield percentages were presented. The authors compared their results with the combined upland red oak group of U.S. Forest Service publication FPL-63 (36) 12 and found that a higher percentage of No. 1 Common and Better lumber was produced from the northern red oak sawlogs, and that a lower percentage of No. BB Common lumber was produced. They warned that value estimates based on FPL-63 may be low for northern red oak. Hanks (12) examined lumber grade yields for 600 subfactory-class logs (those not meeting factory-grade specifications) of various species sawn during Forest Products Laboratory grade yield studies. Regression techniques were used to curve the data, and original subfactory red oak yield data from a previous study by Schroeder (32) were curved and presented. Regression techniques for curving lumber grade yields were used by Hanks and Brisbin (13) for over 600 graded aspen logs in Minnesota. They found that curved results gave better approximation of expected yields than did actual yield summaries. Hanks, at El (14) provided an update to FPL-63 (36) with lumber yield data for 16 hardwood species. Nearly 20,000 graded sawlogs were sawn at over 75 sawmills over a 40-year period. The standard U.S. Forest Service log grade specifications used have remained unchanged since their introduction in 1949 (14). The authors cited log diameter as the single most important variable related to lumber grade yield and log value. A recent study by Wilson 23 El (39) reported lumber recovery and board-foot yield equations for northern l3 thickness at the Ford Forestry Center sawmill over a 25-year period. Yield and quality indexes were shown for five species and one species group designated "other." Red oak was not included. Lag Characteristics Goho and Wysor (10) examined log characteristics of 4,386 hardwood logs measured at 15 sawmills in the eastern U.S. (U.S. Forest Service Region 1). They found that almost two-thirds of the grade 1, 2, and 3 red oak sawlogs (958 log sample) had scaling diameters of 14 inches or less, 75 percent were 15 inches or less, and ninety percent were 18 inches or less. The largest proportion of logs fell into the 12-inch category. The most common log length category was 12 foot, while 10-foot logs were the next most prevalent. Fifty percent of the red oak logs scaled at less than 60 board feet Doyle and 75 percent of the logs had volumes of 100 board feet or less. In 1973, Goho and Martin (11) compared characteristics of logs in Ohio, Kentucky, and Tennessee (U.S. Forest Service Region 2) to those previously studied by Goho and Wysor (10) in U.S. Forest Service Region 1 to see if similar characteristics were common to both regions. Seven species showed significant differences between the regions with respect to log characteristics but, diameter and length distributions for red oak were found to be reliably usable outside of Region 1. In both regions, 75 percent of the logs were 15 inches or less in diameter and 90 percent were 18 14 inches or less. They suggested a need for surveys in other areas of smaller geographic region. PROCEDURE Mill §£l§££i£fl At the time of field data collection, each of the four sawmills selected for study was sawing between one and six million board feet (MMBF) of graded hardwood lumber annually. In a given year, the mills sawed primarily basswood (1111a amariaana), hard and soft maple (Agar aaaaharam and A4 roora and A4 saocharinum), and white and red oak (QHQLQHS alba and Q; roora and Q4 relatina) (40). Both red and black oak, Qaraora and Q4 yelatina, fall under the commercial hardwood designation of "red oak." The mills selected were assumed to represent typical central Michigan hardwood lumber saw- mills. Red oak was an important production component to each of these four mills. Data QQllgfiiiQD EL In: Sasmill At each mill, primarily red oak and some black oak sawlogs were laid out in the mill yards by the operator and numbered sequentially for identification during sawing. Logs were individually graded according to hardwood log grading rules published by the U.S. Forest Service (30) for factory grade logs and scaled using the Doyle log rule. Subfactory- class logs, i.e., construction or local use class (30), were those not meeting factory-grade specifications and were 15 16 jointly designated log grade 4. The graded and scaled logs were debarked and sawn in standard grade- or around-the-log fashion. (F.B. Malcolm (24) described grade sawing in "A simplified procedure for developing grade lumber from hardwood logs.") Each board was numbered by previous log designation and then tallied and graded, either on the green chain or later in piles, by a certified National Hardwood Lumber Association grader. Tally and grade of each board was recorded and yield results by log grade were summarized and tabulated. A total of 404 red oak logs in grades 1, 2, 3, and 4 were studied. Appendix A shows number of logs, log scale, and mill tally for each sawmill in the study. SQLIE Aaalxaia The log and lumber data for each mill were coded on computer input forms and sent to the U.S. Forest Service in Bromall, Pennsylvania for compilation and analysis by a program called SOLVE II. SOLVE II is a computerized sawmill analysis technique developed by the U.S. Forest Service's Northeastern Forest Experiment Station and the Northeastern Area of State and Private Forestry (NA-S&PF) (1). The SOLVE II program provides sawmill managers with an analytical tool to help gain increased yields from logs and to minimize conversion cost. Inputs consist of log and lumber data, prices for sawlogs and products, and basic mill data. Infor- mation provided consists primarily of product yields and break even points for purchasing various species and logs(l). 17 For the purposes of this study, SOLVE II provided grade yield information by diameter for FAS (First and Seconds), FIF (FAS one face), 1C, 2C, 3C lumber and Timbers, as well as diameter and length distributions for the 404 study logs. Also provided were net log volumes and overrun factors for three log rules: International 1/4", Scribner Decimal C, and Doyle. Grade Yield Tabulation Tabulated red oak grade yields were extracted from SOLVE II analyses for the four study mills and were combined by log grade and diameter using a microcomputer. One yield table was produced for each log grade. Aassumptions suggested by Hanks (15) that each mill sawed a normal mix of the 31X lumber grades observed and that all were sawing in around-the-log fashion (grade-sawing) were adapted. These actual green yields were curved using regression techniques to make them suitable for quality index calculations. (McCauley and Mendel (25) found that actual data was too erratic for 01 calculation and that curving was necessary.) The following suggested regression equation was used (12). 1 grade yield : a + bx + cX2 where: X = inches of log diameter inside bark (d.i.b.) A stepwise regression model-selection option produced models with one or both variables (diameter, diameter2 ) that would best predict actual percentage yield data. Residual analysis was performed for each equation to determine whether undesirable trends of deviation occurred in actual versus 18 predicted yields. Some regression analyses were repeated after analysis of residuals indicated that the initial equation did not properly describe actual data. Curved green grade yields were tabulated by log grade and diameter. Regression equations, mean lumber yields, standard error of estimates, and adjusted coefficients of determination for each lumber grade also were tabulated. A total of 22 equations were produced, six for each lumber grade in log grades 1, 2, and 3, and four for log grade 4. No FAS or FIF grade lumber was sawn from grade 4 red oak study logs. Bride BdldLiie 124:4) Qalddladidn A price relative (p.r.) is the ratio of price per thousand board feet for 4/4 lumber of a given grade to that of 4/4 No. 1 Common lumber. Price relatives for red oak were calculated from flarguood Marde Report (22) end-of-the-month weekly price quotes for 4/4 lumber in the Northern Region. The Formula used for calculating the p.r. of each grade was: p.r. = ' (price 4/4 No. 1 Common per MBF) where: lumber grade 1 = grade of interest Twenty quarterly p.r.s were combined to produce an average for the five-year base period July 1978 to April 1983. Price relatives for FAS and FIF included a weekly suggested premium for straight cars of FAS and Selects (FIF) red oak lumber (Appendix B). TIMBER price relatives were l9 estimated using 3A and SD3 Raroaooo Market Report prices after T-testing showed no significant difference between these p.r.s and those calculated from selling price quotes for TIMBERS supplied by operators of the study sawmills. Calculated p.r.s were used to compute log quality indexes and were compared with p.r.s developed by Mendel and Peirsol (28) and Mendel and Smith (27) for the base periods 1970-74, and 1964-68, respectively. Qdaliil Index LQll leduldiidn Curved lumber yields for log grades 1, 2, 3, and 4, and 1978-83 price relatives for red oak were used to calculate quality indexes for each log grade. The formula suggested by Beazley and Herrick (3) was used to generate a log quality index representing the six lumber grades sawn in the study: (1 yield FAS)(p.r. FAS) = X1 (3 yield FIF)(p.r. FIF) : X2 (1 yield 1C)(p.r. IC) = X3 (1 yield 2C)(p.r. 2C) : X4 (3 yield 3C)(p.r. 3C) = X5 (% yield Timbers)(p.r. Ti bers) : X6 QI = XI + X2 + X3 + X4 + X5 + X6 Quality indexes were computed and tabulated by log grade and diameter so that lumber yield trends would be reflected in the 015. 01s reported by diameter may be more useful to sawmills sawing a predominance of any given log diameter. The estimated value of tallied lumber generated at each of the study mills was compared with log-product value estimates calculated from average 015. Net Doyle scale for each log grade and the price of 4/4 No.1 Common red oak 20 product value from average 015: 01 Log Product = QI. x Vj_x P Value 1 where: QIi aveerage quality index log grade 1 Vi = net log volume (Doyle) log grade 1 P a price 4/4 No.1 Common lumber Lumber value per log grade was calculated using percentage yields for each lumber grade and sawmill and September 17, 1983 (22) prices quoted for each grade of 4/4 red oak lumber. Estimated Value of : MBF lumber tally x price/MBF Tallied Lumber grade i 4/4 r.oak Values calculated for each lumber grade were summed, resulting in an estimated value of tallied lumber sawn for each log grade: (MBF FAS)(price FAS) = Y1 (MBF F1F)(price F1F) : Y2 (MBF 1C)(price 1C) = Y3 (MBF 2C)(price 2C) = Y4 (MBF 3C)(Price 3C) - Y5 (MBF Timbers)(price Ti bers) = Y6 Estimated Value of Lumber from = Y1 + Y2 + Y3 + Y4 + Y5 + Y6 Log Grade 1 RESULTS AND DISCUSSION Adtdal Grade Yielda Tables 1 through 4 show actual percentage green lumber grade yields, total board feet, and number of factory and subfactory grade red oak sawlogs, by diameter class. Table 5 compares actual average study yields for each lumber grade with expected yields suggested by U.S. Forest Products Laboratory publication EFL-63. These expected yields were input variables for the SOLVE II analyses. A four-sawmill average was computed for the comparison with actual yields. In table 1, 75 of the 404 study logs were grade 1, yielding 9,959 board feet of grade lumber. For grade 1 sawlogs, average observed yield of No. 1 Common and Better lumber for all diameters was 3.1 percent lower than that suggested by FPL-63; 68.9 percent versus 72.0 percent. Timbers made up 24.9 percent of the yield from grade 1 study logs (Table 5). Of the 16,785 board feet of lumber sawn from 158 grade 2 sawlogs, 49.2 percent was No. '1 Common and Better; 3.7 percent more than predicted by FPL-63. Timbers made up almost 40 percent of the average product yield from grade 2 logs (Tables 2 and 5). One-hundred nineteen grade 3 logs produced 10,033 board feet of lumber, 26.2 percent of which was No. 1 Common or 21 22 omv.m Hem NmN hem.a mmm.H www.mi mmm.m mm NHHmB HoumEMflo Doom UHmon 2H maamu Hmuoa HH< m.e~ e.m m.~ e.ma e.mHI m.mm I I ommuo>< m.mm m.HH o.o o.mm o.m. m.om mam H «N I I I I I I I 0 mm m.mm o.o m.o o.m m.ma v.nv mum m mm m.mm m.H m.H m.oH ~.oa «.me new m Hm m.ma v.H m.v H.ma h.m h.mm omm m om >.ma h.H H.m N.NH m.ma o.vm moa m ma m.mm m.v n.m m.ma m.mH N.ov vmm m ma ~.mm m.v m.m v.m o.m~ m.mm mwma m 5H v.mm m.m N.m o.wa m.ma H.>m has n ma m.mm m.N h.~ v.om o.ba m.Hm mama ma ma H.m~ v.m m.v m.m m.m~ w.mm mmaa NH vH m.om m.a H.m m.ma ~.oH m.vm how oa ma v.vm h.eH >.m «.mm «.5 >.vH mmH N NH muoceHa cm om UH aHm m wmnEdH cmoum Hmsuo< .H manna 23 hmm.m Noo.H mwm mmm.m mom.N HOH.N mmh.mH mmH mHHma HoumEmHa ammo eaaoa an waded Hades HH< n.mm o.m H.m H.mN H.eH o.NH I I ommwo>< m.mm >.n m.m N.NN m.0H o.OH mom N «N m.HN m.h m.N N.Nm v.vN m.m th H mN v.NN N.NH o.o m.vm >.h. ¢.mN Noe N NN m.vv N.m m.H v.mN N.NH m.m mhv N HN H.om m.w m.H m.vN m.mN m.m Nah m ON m.VN m.v m.H m.om h.oN 5.5H HNmH m mH m.mm v.m N.v N.NN m.vH 5.0H NHhH NH mH n.m¢ o.» o.h h.mH h.m N.NH mmON hH hH h.mm m.m m.N m.VN m.qH o.mH OHmH 0H 0H v.mv m.o v.5 m.NH N.NH o.0H HHNN mN mH H.H¢ ¢.m e.m e.NN m.vH N.NH ommH mH «H o.>v h.m o.m m.mH o.MH m.m mmoH NN NH H.nv o.m o.NH v.9N m.n m.m mmm «H NH m.wv H.m H.m n.vH m.NH H.m HHo HH HH m.wm m.v N.NH m.mH m.v o.o NNH m 0H maoceHa om om oH aHm m< m.~A N.m o.o m.eH m.oH o.o mmH H mm I I I I I I I o mm I I I I I I I 0 HA H.mm m.m ~.H m.oe m.mH o.o New H om «.mm m.A o.o e.- e.m A.e mmHH e AH m.mm ~.~ o.o m.oa o.o e.e AmH H mH o.ee m.e H.A m.om H.a e.A Hmm a AH m.ma A.AH o.m m.om o.m A.~ mmmH HH eH A.Ae A.eH m.A m.mH A.m a.e ammH AH mH A.¢A e.OH m.~ N.NH m.m e.H eAm OH «H m.me a.OH m.A o.aH A.m A.H eAm aH AH A.mA o.OH m.m m.m e.o m.H mMOH AH NH «.mm m.~H m.A m.HH e.e A.N HmOH om HH e.me e.HH A.N H.0H a.m H.H meA AH OH muonsHm mm mm UH AHA mam HHHmm mmaH moaoaH meEGH mo .mmumEmHo AHmsuomv mUHmmw mwmma meEDH cmmma mammmmm .um .Um meESZ acHHmom .ucmommm CH .maoH3mm m momma xmo Umm mOM mUHmH> mmnESH Cmmma Hmsmofl .m mHQmB 25 mam.m Amm mm om o o mom.~ mm mHHaa mmumEmHQ comm mmaoa aH AHHmu Hauom HHa o.om m.MH H.m H.m o . o I I ommmo>¢ m.m5 5.HN o.o o.o o o Na H 5H 0.55 v.VH w.m 0.0 o o 5mH H wH N.m5 m.mH m.N m.m o o mmm a mH m.mo m.mN N.a v.m o o Hmv m vH w.¢m v.mH o.o 0.9 o 0 m5 H NH v.55 m.mH m.H o.m o o HNo NH NH m.mm m.m m.v m.N o o mam m HH v.mm o.m o.o 5.H o o HvN m 0H v.Hm m.v v.N v.H o o H5m m a m.Nm H.NH m.m c.o o o eNH m m mommaHm om om oH AHA m momma meESH :mmma ucmommm .um .om meEDZ aCHHmom .ucmommm :H .v momma .maoHBmm mmmHo wmomomwnsm xmo omm mom monH> mmnEsH :mmma Hmsmofl .v mHnt 26 Better. This observed yield was 11.3 percent greater than the expected yield reported in FPL-63. Most was attributed to the exceptionally large proportion of FAS and FIF sawn from grade 3 study logs (Table 5). Fifty-seven percent of the output from grade 3 logs was in the form of Timbers (Table 3). Yield information for subfactory-class red oak logs originally published by Schroeder (32) and later curved via regression techniques by Hanks (12) was used for comparison with actual study yields for subfactory class grade 4 logs. No FAS or FIF lumber was sawn from grade 4 logs. Thus, less than one-half of the expected quantity of No. 1 Common and Better lumber was produced from subfactory logs in the study; 3.1 percent versus 6.9 percent. Timbers comprised 80 percent of the grade 4 log product recovery (Tables 4 and 5). Far lesser quantities of grade 2C and 3C lumber were sawn from study logs than yields for the same grades reported in FPL-63 and by Schroeder. During the study, markets demanded that the sawyer leave No. 3 Common lumber in cant or timber form. In 1981, Timbers brought Michigan sawmill operators an average premium of 30 to 40 dollars over the selling price of No. 3 Common lumber. Figure 1 illustrates the cumulative lumber grade mix for each log grade. Curred Yields and Regression Eddatiena Tables 6 through 9 show curved lumber yields resulting from regression techniques for the factory and subfactory 27 .mm .wnmmo mmmmD ..mum .mxm .mom ammummmsumoz .mmNImz .mmm .mmm .>mmm .mom mam: .maoH ooozommn mmeo mmouomMQSm mow monHm momma meEsH mmmmw .m5mH .m.H .mxcmmN .omHB .cOmHomz .mmommmonmq mposoomm ummmom .meHmm .mmm .mmm .>mmm .mom «om: .mmnEsH ommocmmm mom mmomma aoH ooosommm .mme .Hm um ..H.U .m:asm>H H.mm m a.m m m.mm m.mN o.vH v.0 v.o H.o Nomuommxm v m.om o H.m O o.om N.NH H.m H.m o o om>mmm£o a o.em m m.aH m m.H m.mm A.mm m.mH m.o m.o wouoodxo m m.m5 o N.wN o o.5m m.HH m.v o.mH o.m N.m om>mmmno m 5.mm m m.mv m m.H o.mm ¢.mH m.Hm H.m m.m omuommxm N m.om O N.mv o 5.am o.o H.m H.NN H.v o.NH om>mmmno N o.m~ m o.NA m m.H A.mH A.OH A.em m.m m.mm mopommxo H H.Hm O m.mw O m.VN v.m m.N m.MH 5.w m.mm om>mmmno H moon mam mouuon mam mzm om oN oH AHA mam mmmmm moH GOEEoo N .02 QOEEoo H .02 AwPIonHw momma .maoH mEmm may mew moIHmm Eomm monHm omuommxm nuH3 mHHHEBmm m50m Eomw maoH wosum mow monHm mmnEsH xmo omm Hmsuom mo GOmHmmmEou .m mHQmE IOO‘ QO-I 80- k E 70- ll.) °|~ 60- 8 E 50" m E 40- "a“ 3 30- 1% was as I'—1 28 I TIMBERS 1 3C 2C I l I TIMBERSJ TIMBERS L 06 GRADE Figure 1. Cumulative distribution of lumber grades for red oak sawlogs, by 109 grade 29 class logs studied. Regression equations, mean lumber yields, standard error of estimates, and adjusted coefficients of determination used to generate the tables are displayed in Table 10. The adjusted coefficient of determination, R 2, is a measure of how well the regression curves fit the actual yield data. An R2 of 0.50 means that 50 percent of the variation in the actual data is explained by the regression. For log grade 1, the regression curve for FAS lumber fit the actual data nicely with an R33 of 0.72. Yield data for lumber grades FIF and 1C were the most poorly explained by their regression equations with adjusted R2 s, of 0.00 and 0.17, respectively (Table 6). Regressions for lumber sawn from grade 2 logs were good except for grades FIF and 3C, which had R2 s of 0.17 and 0.00. No. l and No. 2 Common grades demonstrated the highest adjusted R2 s, with 0.66 and 0.59, respectively. Lumber grades FAS and 1C had the highest adjusted R2 for log grade 3, with 0.58 and 0.64, respectively. Regressions for lumber grades FIF and 3C explained little of the actual data's variation, with R2 s of 0.00 and 0.14. Some regressions were poor because of insufficient numbers of logs representing the yield for a particular diameter class. This was especially true for larger logs. For log grade 1, no 23-inch logs, and only one 24-inch log contributed to the regressed yield percentages reported (Table 1). The 23-inch category for log grade 2 was 30 Table 6. Curved green lumber yield for red oak grade 1 sawlogs, in percent. Scaling Percent green lumber grade yields (curved) Diameter, inches FAS FIF 1C 2C BC Timbers 12 19 13 20 3 9 36 13 25 15 18 3 8 31 14 30 16 ‘16 4 6 28 15 34 17 15 4 5 25 16 38 17 14 4 4 23 17 42 17 13 3 3 22 18 45 17 12 3 2 21 19 47 16 12 3 1 21 20 49 15 11 3 l 21 21 50 13 ll 2 l 23 22 51 12 ll 1 1 24 23 51 9 11 l 2 26 24 50 7 ll - 2 30 31 Table 7. Curved green lumber yield for red oak grade 2 sawlogs, in percent. Scaling Percent green lumber grade yields (curved) Diameter, inches FAS FIF 1C 2C 3C Timbers 10 2 9 17 ll 7 54 ll 4 ll 18 10 6 51 12 7 12 19 8 6 48 13 9 13 19 7 6 46 14 11 14 20 6 6 43 15 12 15 21 6 5 41 16 13 16 23 5 5 39 17 13 17 24 4 5 37 18 l4 17 25 3 5 36 19 13 17 26 3 6 35 20 l3 17 27 3 6 34 21 12 17 29 3 6 33 22 10 18 31 2 6 33 23 8 17 32 3 7 33 24 6 17 34 3 7 33 32 Table 8. Curved green lumber yield for red oak grade 3 sawlogs, in percent. Scaling Percent green lumber grade yields Yourved) Diameter, inches FAS FIF 1C 2C 3C Timbers 10 l 5 8 4 10 72 11 l 4 10 5 12 68 12 2 3 12 6 13 64 13 2 3 15 7 13 60 14 3 3 l7 7 13 57 15 4 3 20 6 12 55 16 4 3 23 5 ll 53 17 5 4 26 4 10 51 18 6 5 30 2 8 49 19 7 7 33 - 5 48 Table 9. Curved green lumber yield for red oak subfactory class sawlogs, grade 4, in percent. Scaling Percent green lumber grade yields (curved) Diameter, inches FAS FIF 1C 2C BC Timbers 8 0 0 1 3 6 90 9 0 0 1 3 8 88 10 O O 2 3 9 86 11 O O 2 3 ll 84 12 0 0 3 3 13 81 13 O O 3 2 16 79 14 0 0 4 2 18 76 15 O O 5 2 20 73 16 0 0 5 2 23 7O 17 O O 6 2 26 66 33 Nm. mN.m H.Nm oHoOH.I I H5.wm mmmneme av. No.5 w.NH moamo. I Nm. Um o 0H.N 5.N mAmoo.I I Hm.m UN ON. mN.N w.N m5NNo. I mm.I UH I I I I I I ems I I I I I I was 4 mm. 5N.HH 5.5m om5mH. 5No.mI 5H.mmH mmmnEHB «H. wm.v v.0H m5omN.I mwmm.o mm.HmI Um 5m. em.N 5.v ommHN.I omNm.m om.mmI UN aw. av.0H «.mH mm5mo. I mm.HI UH o 5N.N H.v m5mmH. ONHm.eI wo.mm mHm mm. 5N.N m.m Hemmo. o5vm.I mm.m m mmnEDH cmmE .mcoHumsam conmmmamm .OH mHnme 34 represented by one log. Two logs were in each of the 21-, 22-, and 24-inch diameter categories (Table 2). No 21- or 22-inch grade 3 logs were available for predicting yields in these classes. Only one grade 3 log fell into each of the 18-, 20-, and 23-inch diameter classes (Table 3). Yield distributions for 13-, 16-, and l7-inch grade 4 sawlogs were represented by one log each (Table 4). FIF yields were especially erratic and difficult to explain via regression techniques. Yields did not increase or decrease by diameter in very reliable trends. No. 1 Common lumber yields from log grades 1 and 4 were difficult to describe using regression methods, and No.3 Common yields were poorly explained in log grades 2 and 3. Curved trends for 2C lumber were reliable for all grade logs except subfactory grade 4 logs. Yield data from Timbers and FAS generally produced strongly descriptive equations for all log grades in which they were present. Schroeder (32) and Schroeder and Hanks (31) developed curved yield tables for factory- and subfactory-class red oak logs, which were used for comparison with observed results in Tables 6 through 9. In general, upper grade proportions increased with diameter while lower grades decreased with diameter. Schroeder found that for grade 1 and 2 logs, the curved percentage of FAS lumber continually increased with diameter. FAS lumber sawn from grade 1 logs in this study showed increases up to and a decline after 20 inches d.i.b. (Table 8). 35 Reasons for this decline of FAS yield at upper diameters are speculative. As stated, partial blame may be attributed to the lack of sufficient log information for these diameter classes. However, when northern red oak yield information from U.S. Forest Service publication NE-468 (14) was regressed on diameter for the 13 to 30 inch classes, a similar trend was observed. FAS yield from grade 1 logs reached a maximum of 34.2 percent at 22 inches, then declined to 25.0 percent at 30 inches d.i.b. One could conclude then, that these larger logs are less sound than the smaller logs, accounting for the reduced output of Select and Better lumber. Schroeder's data indicated that grade 3 red oak sawlogs produce about one percent FAS lumber regardless of diameter. Curved FAS yield from grade 3 red oak study logs, however, increased with diameter to a maximum of seven percent. Examination of actual data suggests that the increase drops off abruptly after 19 inches d.i.b. (Table 9). Subfactory grade 4 logs produced No.1 Common yields that increased with diameter. Actual No.2 Common lumber output was much smaller than expected for subfactory class logs. Most log-product output remained in timber form and, as expected, the proportion of timbers decreased with increasing diameter (Table 9). The lumber in these other studies was 4/4 thickness and no timbers were sawn. Timbers made up between 25 and 80 percent of the log-product output observed in this study, r-c-c-II 36 depending on log grade (Table 5). L92 Qharaeieriaiida The majority of the 404 study logs, 39.1 percent, were grade 2 sawlogs. Grade 1 and grade 3 were next most prevalent comprising 18.6 and 29.4 percent of the total, respectively. Grade 4 subfactory class logs made up 12.9 percent of the total. In 1970, Goho and Wysor (10) found that only two percent of Appalachian sawlogs (all species) were subfactory. Individual red oak findings were not reported. As expected, they found that both grade 2 and grade 3 Appalachian red oak sawlogs outnumbered grade 1 logs;i almost 80 percent of the red oak.logs were grade 2 or grade 3 (10). Figure 2 illustrates the distribution of numbers of logs, net log volume (Doyle), and lumber yield tally among the four log grades. The numbers of logs in grades 3 and 4 contributed much less to overall log scale and lumber tally than did the number of logs in grade 1 or 2. Although grade 1 logs made up only 18.6 percent of overall log distribution, they contributed to 26.2 percent of the total net log scale and to 25.1 percent of the final lumber yield tally. Grade 1 logs were larger in diameter than logs from grade 2, 3, or 4. Average diameter for log grade 1 was 16.1 inches d.i.b., while for log grades 2, 3, and 4, average diameters were 15.2, 13.4, and 11.3 inches d.i.b., respectively. Figure 3 illustrates the diameter distribution of the LOG GRADE I LOG GRADE 2 L06 GRADE 3 LOG GRADE 4 37 :1 NO. OF LOGS - NET DOYLE LOG SCALE x\\V LUMBER TALLY PERCENT IO 20 30 4O 50 l j j I 1 fl j Figure 2- Percent distribution of red oak sawlogs, net Doyle scale, and lumber tally, by log grade (404 log sample) 38 .Adeaam moH sows ucmommm :H .maoH3mm xmo omm omomma mOM mmmumEmHo aoH mo GOHuanmumHa .m mmsaHm m N m m w oN aH Hm mm am we 5v me am oN a m mac; mo .02 w.o o.o N.H N.H o.N a.e A.¢ A.A o.m o.eH ¢.HH m.HH N.HH 5.a ¢.m N.N N.H mmowtw HE I I I I I I I m.o N.o o.H N.H N.o o.m o.N m.H N.N N.H a moose I m.o I I m.o m.H N.o o.H A.N N.v m.N m.m N.¢ o.m N.e I I m momma m.o m.o m.o m.o N.H N.N o.m N.¢ o.v A.m A.¢ ¢.m m.m A.N A.o I I N momem m.o o A.o A.o m6 NH mH N.N AH 5m o.m m.N m6 I I I I H255 5zmumma I mom onhzmHmmmHo mm5mzmmuaH moadmmmsoo wmm I mmac.o omeo.o samo.o soHomHtma mmamdmum oa.o os.o om.o sH.H HGIA memsum mos dmamdas mmmIAamH I am.o mm.o AH.H .Hmdda mmmH womHHmz .m GAIAGAH ob eduodflohd eeIAmmH I oe.o mm.o AH.H smoabmoz mmmH .Abodmum eds AmHucdm .m os.o He.o mA.o AH.H eoch I.Hsdds eao eosHs was com mmImmmH m os.o mm.o GA.o eH.H mom I.Hsda< mmmH .Homsoz was mmHamooz .s meIammH I mm.o mm.o mH.H .Hadda oAmH .suHam was Hounds .m sAIOAmH I Hm.o AA.o mo.H N.Hsdda AAmH .Hohmmma mam Homamz .N mmImAmH mv.o Hm.o mm.o AH.H amosomoz mmAH .smsuom .H a madam m dmsmo N datum H madam Hudxmsz Hedda mondd aOH aOH aOH aoq HmcoHamm mmmm mOHmm meEdH mommU aOH an HO mammm>¢ mzm omm mosmm HO .mmmHINmmH Eomm mmHosmm me mo comHmmmEoo "momma aoH an xmo omm mow mmxmoGH wuHHmso .mH mHQmB 48 .mcmommm cH .maoHBmm N momma ocm H momma xmo omm mOM monHm meEsH om>msU .m._.a 3:02. I $5523 mod .N NN ON 2 m: 3 NH 2 O .. wfik Um. . J Itk 3. _ UN OH mammsfim N momma OOH lNBOHSd - 0131A BOVBD .m...o wwIOZ. I mmbm2<_o 004 .m mmsaHm em NA ON 2 m: 3 Na 0 UN 6 on Imw JKHK a. a 8 a 10 wqmmkfk IV a. 10 .I. C... n n I 10 H mocmo .004 9 lNSOUBd - (”BIA 301139 49 .ucmommm CH .maoHBmm v momma omm m momma xmo omm mom monH> mmnEsH om>msU .o mmsaHm .m._.o 3:02. I 55:23 mod .36 3:02. I mmmmzsa 00.. NH 2 .1 NH 2 w ON 2 m: 3 NH OH O p IL P 0 000 ION~ 1 I I 1 T luaouad - O13IA aovao T O 1 08 08 OL 09 OS 017 08 DZ 01 wmmmé: j WIMQIZ m momma .00.. 8 f a momma on: OOI .LNSOHBd - O‘IBIA SOVUD 50 As expected, quality indexes for grade 3 logs and subfactory grade u logs were more stable because of the high proportion of lower-grade lumber they produced. Prices for these lower grades fluctuate less relative to No.1 Common prices than do those for the upper grades. Maxim of Estimated lain: of Tallied Lumber with 9.1. Estimated Lozfimfluct Malia: Table 1” compares the estimated value of tallied lumber produced at each of the study mills with log-product value estimates calculated from average 013. The total value of the 10,000 board feet hypothetical mill tally for all four log grades ranged from $3,92u.1o for Sawmill C to $u,u31.58 for Sawmill D. The average all-grade total for the mills was $u,197.29. The quality index log-product value estimate was $H,005.00 for all log grades, 4.8 percent less than the average estimated value of tallied lumber calculated for the study output. Statistical testing indicated that this difference was not significant, and that the derived QIs provided a good estimate of the average product value sawn from grade 1, 2, 3, and 4 study logs. For each log grade, the log-product value estimate was somewhat lower than that calculated for the lumber tally. Differences between the value of tallied lumber and QI- estimated log-product values ranged from 1.5 percent of the 01 estimate for grade 2 logs, to 15.0 percent of the 01 estimate for grade 3 logs (Table 14). McCauley and Mendel (25) found that derived QIs for grade 3 and subfactory-class 51 .mHHHE boom Ham mo cofiuaoauomflo usmuso mmmum>m co>uomno on» muowfimou mowmum moH ummCOEm poow cumon ooo.oa mnu mo cofiusnfiuumflo mzem .umow cumon ooo.oa mo >Hamu HonEsH Hmoflumnuoaxc m can HHflE sumo um xfiE vamwx momma Hmsuom co comma mfl mafim> uwnESAH m.vu oo.mooq AHo.nHme .mm.nnmv om.mw~ am.nmav mm.amv¢ oa.vmmm mo.maov nm.mave 0.0H mmomuu flHd m.mn mn.nHH Amm.H¢H .mm.vaav mm.m e~.mmfl ~o.m- mm.mma mm.maa mo.m- >.o q o.mHu ov.Hmo Amm.mmm .on.omev n~.mofi Ho.cme mn.mvm om.man om.mmm oo.mon m.~ m m.Hn nm.nvma Amm.mvam .mm.mocav em.HnH oe.enma mo.oo¢~ oo.m~ma oa.mmma mm.moma m.v m m.eu om.nova Amm.-oa .mm.ommHv mo.mm mm.moVH ma.mmma om.mVVH vo.nova oo.oaoa m.~ a oocmummwfia mumEflumm Hm>umucH m x a U m é mHHmB momma w uosooud wocwcflwcoo msHm> Ham: Hafiz Hafiz Haw: moses; moo umoq Ho wma :moz m .uooesq coflaame mo osam> mm: .wcmum mod >3 .wzam> posooua 00H cwumeflummnHO cud: mHHflE3mm xosum ~50w E0um HumnEDH omflaaou mo osfio> boomeflumw mo :Omfiummeoo .va «Home 52 logs underestimated actual value of tallied lumber at the mill by as much as 50 percent. No such pattern was observed for the average values in this study. The difference between actual and QI-estimated log-product value was within nine percent of the 01 estimate for all log grades except grade 3, where the difference was 15 percent. There are some limitations to the above comparison. More valuable information could be gained by using lumber recovery data from a sawmill not included in this study. Such data would be free of any connection with the quality index derivations and thus be more of a test of their accuracy. The comparisons cited do, however, indicate that the calculated average 015 work for these mills. Qxezrua Eaetera; Q; Adiaetee fer ther Les Sealee Quality indexes tabulated by diameter are adequate for estimating the value of individual logs or for log volumes tallied by diameter class. However, when only total log scale for each log grade is known, the proper 01 requires adjustment for overrun, depending on which log rule is used. Average 015 in Table 12 estimate product value for logs scaled by the Doyle rule. When logs are Scaled using International 1/fl" or Scribner Decimal C, these average QIs would overestimate log-product value. Therefore, Table 15 was assembled for 01 users scaling logs by these other rules. Average overrun observed at the four mills for each log grade and log scale are listed in Table 16. 53 Table 15. Red oak quality indexes adjusted for Interna— tional 1/4 and Scribner Decimal C Log Rules. Log Average 01 by Log Rule Grade Doyle Scribner Int. l/4 l 1.19 1.07 0.98 2 0.89 0.79 0.73 3 0.61 0.52 0.47 4 0.46 0.37 0.33 Table 16. Overrun factors by log rule and log grade, in percent. Net Scale Overrun, Percent Grade Doyle Scribner Int. l/4 1 19.4 7.8 -0.2 2 17.0 4.4 -3.6 3 36.5 16.7 5.8 4 54.6 25.8 9.5 CONCLUSIONS AND RECOMMENDATIONS Grade recovery from a sawmill's log input will influence whether the quality index accurately predicts the value of its lumber output. McEwan (27) cites lumber grade recovery as the single most important item affecting costs, and hence profitability in a sawmill. If a mill produces significantly more No.1 Common or Better lumber from a given log than observed in this and other studies (14,36), the red oak 015 will underestimate the final value of tallied lumber. If it produces less No.1 Common and Better for a given grade and diameter of log, the 015 will overestimate the expected value of tallied lumber. Band mills probably will find a need to adjust QIs upward because there is less kerf lost. Operators sawing special products from red oak logs probably will also need to make an adjustment upward because of the high value of special products. The following example illustrates the use of the 01. At a sawmill, 25 MBF of logs were graded and scaled using the Doyle rule: 26 percent were grade 1; 45 percent were grade 2; 22 percent were grade 3; and seven percent were grade 4. The local market price for u/u No.1 Common lumber is $565 per MBF. Using the average quality indexes reported in Table 12 54 55 the QI-estimated log-product value would be: 01 Log- Product MBF Net Price 4/4 No.1 Value Doyle Scale x Common lumber x 01 = Estimate,$ Log grade 1 6.50 x $565 x 1.19 = 4370.28 Log grade 2 11.25 x $565 x 0.89 = 5657.06 Log grade 3 5.50 x $565 x 0.61 = 1895.58 Log grade 4 1.75 x $565 x 0.46 = 454.82 All grades 25.00 - — $12,377.74 Within plus or minus 10 percent, this operator should expect to saw approximately $12,400 worth of lumber, of mixed grades, from the measured logs. Revenues from red oak log- products which are far below that determined for the logs by 01 estimates would indicate that a problem exists in the log buying or log breakdown process. This check on potential log value using a 01 could serve as a simplified indicator of sawmill efficiency. The log quality index has received suprisingly little circulation in industry. With increased application, the concept can be proven valuable to Michigan sawmill operators by helping them keep account of payments for, and revenues from, graded red oak sawlogs. Red oak yields need to be surveyed periodically to monitor how markets for the various grades of lumber affect grade yield mixes being produced at Michigan sawmills, and thus, resulting 015. More yield information for large diameter logs is needed. Also, the red oak QIs presented here need to be tested in individual mill studies in order to better judge their accuracy. Actual value of tallied lumber 56 from study logs, graded and scaled at a mill, should be compared with QI-estimated log product values for the same logs using current prices. Additional information will help us understand the variability of the 01's predictive accuracy, and to gain confidence in their use. LITERATURE CITED LITERATURE CITED ADAMS, E.L. and D.E. DUNMIRE. 1977. SOLVE II: A technique to improve efficiency and solve problems in hardwood sawmills. USDA Forest Serv. Res. Pap. NE-382. Northeastern Forest Exp. Sta., Upper Darby, PA. ANONYMOUS. 1949. Hardwood log grades for standard lumber -- proposals and results. U.S. Forest Prod. Lab. Rep. FPL-1737. Madison, WI. Cited in Vaughan, C.L. et al. 1966. Hardwood log grades for standard lumber. U.S. Forest Serv. Res. Pap. FPL-63. BEAZLEY, R.I. and A.M. HERRICK. 1954. Lumber price relatives: their application in the hardwood quality index. Purdue Univ. Agric. Exp. Sta. Bull. 610. Lafayette, IN. BENTLEY, W.R. and L.L. STREEBY. 1968. Projected quality indexes for estimating hardwood log values. Forestry Res. Note 142. Dept. of Forestry, Univ. of Wisc., Madison, WI. BLYTH, J.E., J. ZOLLNER, and W.B. SMITH. 1982. Michigan sawlog production and sawmill industry, 1978. USDA Forest Serv. Res. Note NC-276. North Central Forest Exp. Sta., St. Paul, MN. BROMLEY, W.S. 1940. The use of log grades in appraising lumber values of selectively cut northern hardwood timber. Pap. Mich. Acad. Sci., Arts, & Letters 26: 135-142. CAMPBELL, R.A. 1952. The advantages of quality indexes in tree grading. pp. 17-20. In: Proceedings of a Symposium on a Standard Hardwood Quality Index. Purdue Univ., Lafayette, IN. CUMMINS, W.R. 1952. Historical and statistical stability of price ratios. pp. 34-50. In: Proceedings of 3 Symposium on a Standard Hardwood Quality Index. Purdue Univ., Lafayette, IN. ELLERTSON, B.W. and P. LANE. 1953. Lumber price ratios for computing quality index of Tennessee Valley hardwoods. Tech. Note No. 15. T.V.A. Division of Forestry Relations. Cited in Herrick, A.M. 1956. The quality index in hardwood sawtimber management. Agric. Exp. Sta. Bull. 632. 57 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 58 GOHO, C.D. and P.S. WYSOR. 1970. Characteristics of factory-grade hardwood logs delivered to Appalachian sawmills. USDA Forest Serv. Res. Pap. NE-166. Northeastern For. Exp. Sta., Upper Darby, PA. GOHO, C.D. and A.J. MARTIN. 1973. Sawlog sizes: a comparison in two Appalachian regions. USDA Forest Serv. Res. Note NE-160. Northeastern For. Exp. Sta., Upper Darby, PA. HANKS, L.F. 1973. Green lumber grade yields for subfactory class hardwood logs. USDA Forest Serv. Res. Pap. NE-256. Northeastern For. Exp. Sta., Upper Darby, PA. HANKS, L.F. and R.L. BRISBIN. 1978. Lumber grade yields for graded aspen logs and trees. USDA For. Serv. Res. Pap. NE-423. Northeastern For. Exp. Sta., Bromall, PA. HANKS, L.F., G.L. GAMMON, R.L. BRISBIN, and E.D. RAST. 1980. Hardwood log grades and lumber grade yields for factory lumber logs. U.S. Forest Serv. Res. Pap. NE-468. Northeastern For. Exp. Sta., Bromall, PA. HANKS, L.F. U.S. Forest Service, Regional Office, Milwaukee, WI. Private communication, August 1983. HERRICK, A.M. 1946. Grade yields and overrun from Indiana hardwood sawlogs. Purdue Univ. Agric. Exp. Sta. Bull. 519. Lafayette, IN. HERRICK, A.M. 1956. The quality index in hardwood sawtimber management. Agric. Exp. Sta. Bull. 632. Purdue Univ., Lafayette, IN. HUBER, H.A., BOZAAN, D., and L. DRAPER. 1983. The internal cost of producing sawmill residues in Michigan. Agric. Exp. Sta. Res. Rep. 451. Michigan State Univ., E. Lansing, MI. HUYLER, N.K. 1978. Live-sawing: a way to increase lumber grade yield and mill profits. USDA Forest Serv. Res. Pap. NE-305. Northeastern For. Exp. Sta., Upper Darby, PA. JOHNSON, J.A. and J.W. METEER. 1979. Application of log and tree quality index factors for hardwood sawtimber. Ford Forestry Center, Michigan Tech. Univ., Houghton, MI. Presented to the Northern Hardwood & Pine Manuf. Assoc., Inc. Annual Seminar. KRAMER, P.R. 1952. A basis for hardwood log quality indexes. pp. 56-59. In: Proceedings of a Symposium on a Standard Hardwood Quality Index. Purdue Univ., Lafayette, IN. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 59 LEMSKY, ABE. 1978-83. The Hardwood Market Report. Memphis, Tenn. LOCKHARD, C.R. p.61. In: Proceedings of a Symposium on a Standard Hardwood Quality Index. Purdue Univ., Lafayette, IN. MALCOLM, F.B. 1965. A simplified procedure for developing grade lumber from hardwood logs. U.S. Forest Serv. Res. Note FPL-98. For. Prod. Lab., Madison, WI. McCAULEY, O.D. and J.J. MENDEL. 1969. Adjusting quality index log values to represent local and regional commercial sawlog product values. USDA Forest Serv. Res. Pap. NE-149. Northeastern For. Exp. Sta, Upper Darby, PA. McEWAN, T.K. 1974. Analysis of the effect of the lumber recovery factor (LRF) on sawmill costs. For. Prod. J. 24: 17-190 MENDEL, J.J. and W.H. SMITH. 1970. Quality index tables for some eastern hardwood species. USDA For. Serv. Res. Pap. NE-167. Northeastern Forest Exp. Sta., Upper Darby, PA. MENDEL, J.J. and M.K. PEIRSOL. 1977. Quality index tables for some eastern hardwood species based on lumber prices from 1970 to 1974. USDA Forest Serv. Res. Pap. NE-370. Northeastern For. Exp. Sta., Upper Darby, PA. NORTHERN HARDWOOD and PINE MANUFACTURERS ASSOCIATION. 1968. Hardwood sawlog grades for standard lumber. RAST, 5.0., D.L. SONDERMAN, and G.L. GAMMON. 1973. A guide to hardwood log grading (revised 1979). USDA Forest Serv. Gen. Tech. Rep. NE-1. Northeastern For. Exp. Sta., Upper Darby, PA. SCHROEDER, J.G. and L.F. HANKS. 1967. Lumber grade-yields for factory—grade northern red oak sawlogs. U.S. Forest Serv. Res. Note NE-65. Northeastern For. Exp. Sta., Upper Darby, PA. SCHROEDER, J.G. 1968. Lumber grade yields for subfactory class red oak logs. USDA Forest Serv. Res. Pap. NE-114. Northeastern Forest Exp. Sta., Upper Darby, PA. Cited in Hanks, L.F. 1973. Green lumber grade yields for subfac- tory class hardwood logs. USDA Forest Serv. Res. Note NE-256. STREEBY, L.L. and W.R. BENTLEY. 1968. QUINDEX: A program to estimate and project quality indexes for log grades. Forestry Res. Note 137, Univ. of Wisconsin, Madison, WI. 3“. 35. 36. 37. 38. 39. 40. 60 STUMP, W. 1975. Relationship of hardwood log and lumber grades. South. Lumberman 231 (2867): 11. TRIMBLE, C.R., Jr. and J.J. MENDEL. 1969. The rate of value increase for northern red oak, white oak, and chestnut oak. USDA Forest Serv. Res. Pap. NE-129. Northeastern For. Exp. Sta., Upper Darby, PA. VAUGHAN, C.L., A.C. WOLLIN, K.A. McDONALD, and E.H. BULGRIN. 1966. Hardwood log grades for standard lumber. U.S. Forest Serv. Res. Pap. FPL-63. Forest Prod. Lab., Madison, WI. WALLACE, O.P. 1952. The use of quality indexes in designing log grades. pp. 11-16. In: Proceedings of a Symposium on a Standard Hardwood Quality Index. Purdue Univ., Lafayette, IN. WHITE, H.G. 1937. An introduction to the study of timber quality in Lake States hardwoods. Pap. Mich. Acad. Sci., Arts, & Letters 28: 339-350. WILSON, D.W., J.S. KUCAB, J.S. METEER, and J.A. JOHNSON. 1982. Lumber recovery and board foot yield equations for northern hardwoods. Ford Forestry Center Res. Note 30. Michigan Tech. Univ., Houghton, MI. ZOLLNER, J. 1983. Michigan Directory of wood-using companies. Forest Management Div., Mich. Dept. Nat. Res., Lansing, MI. APPENDIX A NUMBER OF LOGS, NET DOYLE LOG SCALE, MBF LUMBER TALLY, AND AND PERCENT OVERRUN FOR EACH SAWMILL, BY LOG GRADE 61 m.¢m mvm.mm mom.am vow Hobos ccmuo I I I I Q v.w> omo.H Hom.o ma 0 H.¢v mom.a mm~.H em m I I I I 4 v H.mm vmv.m man.m mm a 5.Hv mma.m omm.a Hm U m.>¢ mmo.m hmh.a om m m.om mom.H omm.H mm < m m.oH omo.o mH~.m av a H.mm mmm.m hem.m me u m.mI vmo.m nmm.m Hm m o.mm mno.v mmm.m ow < N m.om mmm.m mmm.m mm a o.hm mmm.o mmm.o m U m.0I Hem.a mmm.a ea m n.mm mmm.v mmm.m om d H asuum>o madme Ammzv mamom mos mmoq Hafiz condo unwound nonficq mm: maxoo umz mo .02 mom mwmuw 004 ms .HHH53mm comm now snuum>o Dcmoumm cam .maame Hwnfidq mm: .mamom moq maaoo uwz .mmoq mo HmQEsz fl xHDmem¢ APPENDIX B QUARTERLY 4/4 RED OAK LUMBER PRICES AND FAS/SEL PREMIUMS, 1978-83 APPENDIX B Quaterly 4/4 Red Oak Lumber Prices and FAS/SEL Premiums, 1978-83 1 4 Price per Grade, $/MBF Year and Qtr Date FAS SEL2 1c 2c 3c TMB3 Prem 3 1978 July 29 480 460 385 195 150 175 150 4 Oct 28 495 475 395 200 155 180 150 l 1979 Jan 20 505 485 415 215 160 195 200 2 Apr 28 505 485 415 215 170 195 200 3 July 28 505 485 415 215 180 195 150 4 Oct 27 505 485 405 215 180 195 150 1 1980 Jan 26 505 485 390 205 172 185 150 2 Apr 26 505 485 355 195 160 175 150 3 July 26 505 485 345 185 145 165 115 4 Oct 25 505 485 325 170 135 174 90 1 1981 Jan 24 510 490 330 170 140 150 100 2 Apr 18 510 490 360 173 145 153 115 3 July 25 520 500 380 173 145 153 135 4 Oct 31 520 500 385 173 140 153 110 l 1982 Jan 30 520 500 385 170 137 150 75 2 Apr 24 520 500 390 170 137 150 80 3 July 31 540 520 390 185 137 165 90 4 Oct 30 540 520 400 190 137 170 90 l 1983 Jan 22 540 520 410 215 137 175 90 2 Apr 16 580 560 460 235 137 195 125 Mean (§) 516 496 387 193 150 164 126 lLemsky, Abe. 1978-83. The Hardwood Market Report. 2 Memphis, Tenn. Select prices used for F1F prices 33A/SD3 prices used to represent Timbers prices 4Premium paid used for straight cars of FAS and Selects 62 APPENDIX C QUARTERLY RED OAK PRICE RELATIVES, 1978-83 Quarterly Red Oak Price Relatives, APPENDIX C 1978-83 Price Relatives (p.r.) Year and Qtr Date FAsl FlFl 1c 2c 30 TMB 3 1978 July 29 1.64 1.58 1.0 .506 .390 .454 4 Oct 28 1.63 1.58 1.0 .506 .392 .462 1 1979 Jan 20 1.70 1.65 1.0 .518 .386 .470 2 Apr 28 1.70 1.65 1.0 .518 .410 .470 3 July 28 1.58 1.53 1.0 .518 .434 .470 4 Oct 27 1.62 1.57 1.0 .531 .444 .482 1 1980 Jan 26 1.68 1.63 1 0 .526 .441 .474 2 Apr 26 1.84 1.79 1.0 .549 .451 .493 3 July 26 1.80 1.74 1.0 .536 .420 .478 4 Oct 25 1.83 1.77 1.0 .523 .415 .535 1 1981 Jan 24 1.85 1.79 1.0 .515 .424 .454 2 Apr 18 1.74 1.68 1.0 .481 .403 .425 3 July 25 1.72 1.67 1.0 .455 .382 .403 4 Oct 31 1.64 1.58 1.0 .449 .364 .397 1 1982 Jan 30 1.55 1.49 1.0 .442 .356 .390 2 Apr 24 1.54 1.49 1.0 .436 .351 .385 3 July 31 1.62 1.56 1.0 .474 .351 .423 4 Oct 30 1.58 1.53 1.0 .475 .342 .425 1 1983 Jan 22 1.54 1.49 1.0 .524 .334 .427 2 Apr 16 1.53 1.49 1.0 .511 .298 .424 Standard Dev. (5) .1050 .1014 0 .0337 .0418 .0390 Mean (3) 1.66 1.61 1.0 .500 .389 .447 1Price relatives for FAS and F1F include weekly suggested premiums for straight cars FAS and SEL. 2TMB — Timber price relatives are calculated from 3A/ SD3 prices. 63 Statistical testing showed that there were no differences between Observed Timber p.r.s and that for 3A/SD3. APPENDIX D FOREST SERVICE STANDARD GRADES FOR HARDWOOD FACTORY LUMBER LOGS APPENDIX D Forest Service Standard Grades for Hardwood Factory Lumber Logsa . Log grades Grading Factors F1 F2 F3 Position in tree Eggs 2::er Butts & uppers 13;:er Scaling diameter, inches 13-15" 16-19 20+ 11+" 12+ 8+ Length without trim, feet 10+ 10+ 8-9 10-11 12+ 8+ Required Min. length, feet 7 5 3 3 3 3 3 2 clear cuttings“ No of each of 3 Max. number 2 2 2 2 2 2 3 limit best faces“ Nfliln. [iropogtion o og engt , / . , / required in clear 5 6 5 6 5/6 2 3 3’4 2 3 2/3 1/2 cutting Maxim&um 1: ]For lcflgs with sweep . croo esst an 14 of ,- ea allowance end in souhd 15 $2 30 (7‘ 50 fl defects For log: with f more t an 4 o , ,2. ,.. end in sound 10 9‘ 20 "‘ 35 '" defects Maximum scaling deduction 4O ‘I’c ‘ 50 7c 3 50 C} End defect: ' From USDA Forest Service Research Pa See special instructions (page 18) r FPL-63 (l3). " Ash and basswood butts can he 12 inches i they otherwise meet requirements for small #l's. " Ten-inch logs of all species can be #2 if they otherwise meet re "A clear cuttin is a portion of a face. extendin the width of the ' A face is V. o the surface of the log as divide leng tlm‘Ise. ' Otherwise #1 logs with 4] 450% deductions can he # 2. ‘ Otherwise #2 logs With 51-60% deductions can be # 3. Reprinted from Rast and others (30), p.11 64 uirements for small #l's. ace. that is free of defects.