Short-rotation forestry of birch,maple, poplar and willow in Flanders (Belgium) I—Biomass production after 4 years of tree growth

During the last three decades, oil crises, agricultural surpluses and global climate change enhanced the interest in short-rotation forestry (SRF). In this study, the biomass production of birch (Betula pendula Roth), maple (Acer pseudoplatanus L.—Tintigny), poplar (Populus trichocarpa × deltoides—Hoogvorst) and willow (Salix viminalis—Orm) growing under a short-rotation (SR) management system were compared after a 4 years period. The plantation was established on former agricultural land. The sandy soil had a mean pH of 4.5 and a mean carbon content of 1.0%. Survival rates after 4 years were 75.8%, 96.8%, 86.3% and 97.6% for birch, maple, poplar and willow, respectively. The mean actual annual biomass production for these four species amounted to 2.6, 1.2, 3.5 and 3.4 t DM ha⁻¹ yr⁻¹, respectively. The large variation in biomass production at the different plots of the plantation could not be explained by the measured soil parameters. Biomass production results found here were in the lower range of values reported in literature. However, in contrast to most other studies, no weed control, fertilisation or irrigation was applied in this experiment. As marginal agricultural soils are suboptimal for the growth of poplar and willow, birch can be considered as a very interesting alternative for the establishment of SR plantations in Flanders.     

Estimation of chemical traits in poplar short-rotation coppice at stand level

In order to improve qualitative traits of harvestable biomass and enhance its conversion into second generation biofuels (e.g. bioethanol), much attention should be paid to manage woody-energy plantations. This work represents an attempt to estimate chemical composition of biomass at stand level in poplar plantations. Based on the relationship between chemical traits of stem cross-sections and the corresponding distribution of diameter classes within the whole plantations, three different harvesting cycles were compared in terms of biomass yield and chemical composition. Under 2-year rotation, the stand showed the lower biomass annual yield (11.7 tDM ha^?1 y^?1) and the lower cellulose (42.5%) and the higher lignin (22%) proportion. On the contrary, under the 4-year cutting cycle, annual yield was not only higher (18.4 tDM ha^?1 y^?1), but the biomass also presented the highest cellulose (51.6%) and the lower lignin (19%) share. These results suggest that different management practices, including the cutting cycle, may affect not only yields but also qualitative traits of harvestable biomass of poplar short-rotation coppice.     

Switchgrass production for the upper southeastern USA: Influence of cultivar and cutting frequency on biomass yields

Switchgrass (Panicum virgatum L.) is considered a good biofuels feedstock candidate. However, limited information is available on its productivity and harvest management in the upper southeastern USA. Our objective was to examine production potential of upland and lowland switchgrass cultivars in response to one- or two-cut management across the region. Upland (‘Cave-in-Rock’ and ‘Shelter’) and lowland (‘Alamo’ and ‘Kanlow’) cultivars were harvested for 3 yr under one- or two-cut management at eight sites in five states (North Carolina, Kentucky, Tennessee, Virginia, and West Virginia). Across all sites, years, and cutting managements, upland cultivars yielded 12.6 vs. 15.8 Mg ha⁻¹ for lowland cultivars. Both cultivars yielded more on average with two harvests rather than one, but the effect was greater for upland cultivars (36% more biomass), while lowland cultivars yielded only 8% more biomass with two harvests. Tiller densities were higher for Alamo (lowland) than for Cave-in-Rock (upland) and higher with two-cut than with one-cut management. Early season production of Alamo (a cultivar of southern origin) appeared sensitive to low temperatures. Weak linear responses to precipitation were observed for first-cut biomass, but none was observed for summer precipitation. Lowland switchgrass cultivars appear better suited to biomass production in the upper southeastern USA, due to their greater productivity. Two vs. one cutting per year may be of less advantage for biomass yield with lowland cultivars in this region. However, if upland cultivars are used, two harvests may be of benefit dependent upon production costs and feedstock quality.     

Biomass production from traditional coppice management in northern Italy

Traditional coppice stands cover millions of hectares throughout Europe and offer large amounts of biomass. The study analyzed 10 commercial coppice harvesting operations in northwestern Italy, where modern machines were deployed. Removals, prices, work, revenues and costs were carefully determined. Firewood was the main product, representing between 70% and 100% of the total product mass and value. Traditional coppice stands often yield over 200 m³ of energy biomass per hectare, at the time of cut. Cable yarding operations were better organized than ground-based operations, which explained why they incurred the same harvesting cost, despite the more challenging site conditions under which they were deployed. Mean harvesting cost was 45 € m⁻³, of which about 10% was needed for felling, 70% for extraction and processing, and the remaining 20% for loading and transportation. All operations accrued some profit, which varied between 13 and 43 € m⁻³ or between 1600 and 8600 € ha⁻¹, depending on operational efficiency, value recovery and stand yield.     

Size reduction of high- and low-moisture corn stalks by linear knife grid system

High- and low-moisture corn stalks were tested using a linear knife grid size reduction device developed for first-stage size reduction. The device was used in conjunction with a universal test machine that quantified shearing stress and energy characteristics for forcing a bed of corn stalks through a grid of sharp knives. No published engineering performance data for corn stover with similar devices are available to optimize performance; however, commercial knife grid systems exist for forage size reduction. From the force–displacement data, mean and maximum ultimate shear stresses, cumulative and peak mass-based cutting energies for corn stalks, and mean new surface area-based cutting energies were determined from 4–5 refill runs at two moisture contents (78.8% and 11.3% wet basis), three knife grid spacings (25.4, 50.8, and 101.6 mm), and three bed depths (50.8, 101.6, and 152.4 mm). In general, the results indicated that peak failure load, ultimate shear stress, and cutting energy values varied directly with bed depth and inversely with knife grid spacing. Mean separation analysis established that high- and low-moisture conditions and bed depths ≥ 101.6 mm did not differ significantly (P < 0.05) for ultimate stress and cutting energy values, but knife grid spacing were significantly different. Linear knife grid cutting energy requirements for both moisture conditions of corn stalks were much smaller than reported cutting energy requirements. Ultimate shear stress and cutting energy results of this research should aid the engineering design of commercial scale linear knife gird size reduction equipment for various biomass feedstocks.     

Above-ground biomass production and allometric relations of Eucalyptus globulus Labill. coppice plantations along a chronosequence in the central highlands of Ethiopia

Eucalyptus plantations are extensively managed for wood production in the central highlands of Ethiopia. Nevertheless, little is known about their biomass (dry matter) production, partitioning and dynamics over time. Data from 10 different Eucalyptus globulus stands, with a plantation age ranging from 11 to 60 years and with a coppice-shoot age ranging from 1 to 9 years were collected and analyzed. Above-ground tree biomass of 7–10 sampled trees per stand was determined destructively. Dry weights of tree components (W_c; leaves, twigs, branches, stembark, and stemwood) and total above-ground biomass (W_a) were estimated as a function of diameter above stump (D), tree height (H) and a combination of these. The best fits were obtained, using combinations of D and H. When only one explanatory variable was used, D performed better than H. Total above-ground biomass was linearly related to coppice-shoot age. In contrast a negative relation was observed between the above-ground biomass production and total plantation age (number of cutting cycles). Total above-ground biomass increased from 11 t ha^?1 at a stand age of 1 year to 153 t ha^?1 at 9 years. The highest dry weight was allocated to stemwood and decreased in the following order: stemwood > leaves > stembark > twigs > branches. The equations developed in this study to estimate biomass components can be applied to other Eucalyptus plantations under the assumption that the populations being studied are similar with regard to density and tree size to those for which the relationships were developed.     

Biomass accumulation in rapidly growing loblolly pine and sweetgum

Loblolly pine (Pinus taeda) and sweetgum (Liquidambar styraciflua) trees, growing in International Paper Company's study of intensive management on marginal agricultural land near Bainbridge GA, were destructively sampled at the end of the sixth growing season. All trees were single family blocks of genetically superior trees planted 2.5 m apart on sub-soiled rows 3.6 m apart and grown with complete competition control. Management treatments were: control, irrigation, irrigation plus fertilization, and irrigation plus fertilization plus pest control. Tree measures were basal diameter, DBH, height of live crown, diameter at base of live crown, and total height. Twenty trees of each species were destructively sampled. Stems were sectioned at 1 m intervals, stem diameter determined at each end and sections were weighed green. Branches were removed and height, basal diameter, and length were measured on each branch. Branches were separated into foliated and unfoliated segments and weighed green. A stem disk and branch from each meter were returned to the lab to determine dry weight: green weight ratio. Foliated limb: foliage ratios were also determined from sub-sampled branches. Intensive culture resulted in larger growth differences for sweetgum (most intensive treatment 9.5 m tall, 13.1 cm DBH; control trees 5.0 m tall, 6.3 cm DBH) than in pine (most intensive treatment 10.3 m tall, 17.7 cm DBH; control, 7.6 m tall, 13.4 cm DBH). The pipe model of tree development explained dimensions of the upper 5 m of crown with leaf biomass highly correlated to branch basal area (r² from 0.697 to 0.947). There was a constant ratio of leaf biomass to branch basal area (50 gm/cm² for pine, 30 gm/cm² for sweetgum). We also found a constant ratio of bole basal area to cumulative branch basal area throughout the crowns. Rapidly growing pines produced about 49 Mg ha⁻¹ of stem biomass, 11 Mg ha⁻¹ of dead branch biomass, and 17 Mg ha⁻¹ of unfoliated branch biomass at the end of six years.     

Biomass and volume yield after 6 years in multiclonal hybrid poplar riparian buffer strips

In this paper the potential of five hybrid poplar clones (Populus spp.) to provide biomass and wood volume in the riparian zone is assessed in four agroecosystems of southern Quebec (Canada). For all variables measured, significant Site effects were detected. Survival, biomass yield and volume yield were highest at the Bromptonville site. After 6 years of growth, total aboveground biomass production (stems + branches + leaves) reached 112.8 tDM/ha and total leafless biomass production (stems + branches) reached 101.1 tDM/ha at this site, while stem wood volume attained 237.5 m³/ha. Yields as low as 14.2 tDM/ha for total biomass and 24.8 m³/ha for total stem volume were also observed at the Magog site. Highest yields were obtained on the most fertile sites, particularly in terms of NO₃ supply rate. Mean stem volume per tree was highly correlated with NO₃ supply rate in soils (R² = 0.58, p < 0.001). Clone effects were also detected for most of the variables measured. Total aboveground biomass and total stem volume production were high for clone 3729 (Populus nigra × P. maximowiczii) (73.1 tDM/ha and 134.2 m³/ha), although not statistically different from clone 915311 (P. maximowiczii × P. balsamifera). However, mean whole-tree biomass (including leaves) was significantly higher for clone 3729 (38.8 kgDM/tree). Multifunctional agroforestry systems such as hybrid poplar riparian buffer strips are among the most sustainable ways to produce a high amount of biomass and wood in a short time period, while contributing to alleviate environmental problems such as agricultural non-point source pollution.     

Aboveground tree biomass in a recovering tropical sal (Shorea robusta Gaertn. f.) forest of Eastern Ghats,India

Aboveground biomass of individual tree species by component and total biomass per unit area for four different stages of a recovering tropical dry deciduous forest stands, dominated by sal (Shorea robusta Gaertn. f.) of the Eastern Ghats, India were investigated during 2001–2002. Different periods of recovering (2, 4, 6, and 10-year) forest stands (84°13′E, 20°29′N) were selected in the Kandhamal district of Orissa, India and sample trees of all species were harvested. Tree species diversity was 23, 23, 21 and 22 in 2, 4, 6, and 10-year recovering stands, respectively. Species-wise Ixora pavetta showed the highest biomass in 2 and 4-year stands while Shorea robusta in 6 and 10-year stands. Component-wise, in all species, bole–wood contribution ranged between 22.6% and 60.9%. Aboveground tree biomass, in all the stands, was dominated by Shorea robusta, which ranged between 12.68 and 231.91 Mg ha⁻¹. Total aboveground tree biomass was 30.12, 49.21, 107.54 and 261.08 Mg ha⁻¹ in 2, 4, 6 and 10-year stands, respectively.     

Quantification of the residual biomass obtained from pruning of trees in Mediterranean almond groves

This research quantified the available residual biomass obtained from pruning almond trees. The additional biomass quantified could be used as a source of energy or as raw material for the wood industry and would provide additional income for fruit producers and also a more sustainable system. Several factors were analyzed: Variety, aim of the pruning, age of the plants, size of the plantation, crop yield and irrigation. Regression models were also calculated to predict the weight of dry biomass obtained per tree and tonnes of dry biomass obtained per hectare according to the significant factors. These equations could implement logistic planning as the Borvemar model, which defines a logistics network for supplying bioenergy systems. Almond tree varieties were classified into three groups: a first group with high residual biomass productivity (average yield 12.6 kg dry biomass/tree), a second group with low productivity (average yield 4.5 kg dry biomass/tree) and a transition group with a intermediate biomass yield of 7 kg dry biomass/tree. This means that in Mediterranean areas the residual biomass from almond pruning reaches an average 1.34 t/ha annual.     

Estimation of woody biomass production from a short-rotation bio-energy system in semi-arid Australia

Short, 3–5 year, rotations of trees have been proposed as a method of regaining hydrological control of dryland farming systems (300–600 mm annual rainfall) in southern Australia and thus alleviating salinization of land and water. At the termination of the rotation, the trees will be removed and used as a bioenergy feedstock. In the absence of any tree growth data in this region, allometric relationships were developed for three prospective short-rotation species (Eucalyptus globulus, Eucalyptus occidentalis and Pinus radiata), for 3-year-old trees, at a site with a mean annual rainfall of 365 mm. Equations that related stem diameter over bark at 10 cm (D₁₀) and tree height (ht) to total tree biomass (above and below ground), leaves, stems (stemwood and bark) and roots were developed, by combining data from different planting densities (500, 1000, 2000 and 4000 stem ha⁻¹) and landscape positions (upper-slope, mid-slope and lower-slope).Mean oven-dry yields of the three species, in the high planting density treatment were not significantly different and ranged from 12 to 14 t ha⁻¹ (3 years)⁻¹. There were consistent increases in biomass yield with planting density, with this generally being greatest with the 4000 stem ha⁻¹ treatment. There were marked differences in productivity with slope position. For E. globulus and E. occidentalis the best yields were obtained in lower landscape positions with initial planting densities of 4000 stem ha⁻¹, with 16.6 and 22.2 t ha⁻¹ (3 years)⁻¹, total biomass produced, respectively. The best yield of P. radiata was 15.4 t ha⁻¹ (3 years)⁻¹ from an initial planting density of 4000 stem ha⁻¹ in an upper landscape position. These differences partly reflected site hydrology, with water accumulating in downslope positions. Partitioning of tree components was variable between species, with root:shoot (R:S) ratio being significantly (P<0.0001) higher for E. occidentalis (0.5) compared with the other two species (0.3). Results suggest that biomass productivity can be optimized in this region by using high initial planting densities and recognizing the interaction of different species with site hydrology.     

Above-ground biomass production and nutrient accumulation in young stands of silver birch on abandoned agricultural land

During the last decade, more than 400 000 ha of agricultural land was abandoned in Estonia. Such areas are often characterized by rapid natural afforestation with silver birch, which has led to an increase both in the woodland area and in the area of silver birch stands. However, many bioenergetic aspects related to birch stands growing on arable land are still poorly understood. The main aim of the present study was to investigate the above-ground biomass production, nutrient (NPK) accumulation, and foliar characteristics of young silver birch stands on abandoned agricultural land. Five 8-year-old stands of silver birch growing on different soil types were included in the study.The density of the studied stands varied from 3060 to 36 200 trees per ha and their above-ground biomass varied from 6.0 to 22.9 t DM ha⁻¹. The largest share in the above-ground biomass of the birches (59–80%) was from the stems. The mean stem mass of the birches ranged from 0.29 to 1.79 kg, and the mean total above-ground biomass ranged from 0.36 to 3.03 kg. The leaf area index for the studied stands varied from 1.21 to 4.64 m² m⁻², being the highest for the stand of medium density. Mean single leaf area varied from 9.4±0.2 to 15.4±0.3 cm², leaf weight per area varied from 61.1±0.4 to 77.5±0.5 g m², and specific leaf area varied from 13.2±0.1 to 16.8±0.1 m² kg⁻¹. However, no significant differences were found between stand density and the foliar characteristics. There was a strong positive correlation between soil nitrogen concentration and leaf nitrogen concentration (R=0.92); regarding phosphorus concentration, the corresponding correlation was weak (R=0.52) and regarding potassium concentration, no significant correlation was found. The amount of nitrogen accumulated in the above-ground part of the silver birch stands varied between 42.4 and 145.8 kg ha⁻¹, the amount of phosphorus, between 5.9 and 27.9 kg ha⁻¹, and the amount of potassium, between 7.2 and 78.6 kg ha⁻¹. The N:P:K ratios for the foliage were comparable. It is evident that the proportion of nitrogen and phosphorus are close to optimum, while the N:K ratio was lower than optimum value in all cases.     

Long-term yield potential of switchgrass-for-biofuel systems

Limited information is available regarding biomass production potential of long-term (>5- yr-old) switchgrass (Panicum virgatum L.) stands. Variables of interest in biomass production systems include cultivar selection, site/environment effects, and the impacts of fertility and harvest management on productivity and stand life. We studied biomass production of two upland and two lowland cultivars under two different managements at eight sites in the upper southeastern USA during 1999–2001. (Sites had been planted in 1992 and continuously managed for biomass production.) Switchgrass plots under lower-input management received 50 kg N ha⁻¹ yr⁻¹ and were harvested once, at the end of the season. Plots under higher-input management received 100 kg N ha⁻¹ (in two applications) and were harvested twice, in midsummer and at the end of the season. Management effects on yield, N removal, and stand density were evaluated. Annual biomass production across years, sites, cultivars, and managements averaged 14.2 Mg ha⁻¹. Across years and sites, a large (28%) yield response to increased inputs was observed for upland cultivars; but the potential value of higher-input management for lowland cultivars was masked overall by large site×management interactions. Nitrogen removal was greater under the higher-input system largely due to greater N concentrations in the midsummer harvests. Management recommendations (cultivar, fertilization, and harvest frequency), ideally, should be site and cultivar dependent, given the variable responses reported here.     

Yield and canopy characteristics of switchgrass (Panicum virgatum L.) as influenced by cutting management

Perennial warm season grasses, such as switchgrass (Panicum virgatum L.), complement cool-season species by filling the period of low biomass production called summer-slump. Although switchgrass is widespread in the Great Plains of the USA, there is very little information available about its management in mountain regions. A field experiment was conducted in Blacksburg, VA, USA, to determine the influence of dates of first and regrowth harvest, and cutting height on the canopy structure of the winter hardy cv. ‘Pathfinder’. The number, height and weight of tillers, yield distribution, light penetration through the canopy, leafiness, specific leaf area and leaf area index were investigated. Harvested biomass yield increased as date of first harvest was delayed throughout June and cutting height was reduced from 300 mm to 200 mm, but the yield of regrowth declined accordingly. This decline was caused by the removal of an increasing percentage of apical meristems and increasing damage to the crop canopy, in particular the loss of a greater proportion of photosynthetically active leaf area. A lower cutting height also reduced the amount of biomass produced in the following year. Cutting at 300 mm in early June can maximise the amount of high quality (defined as high concentrations of in vitro digestible dry matter and crude protein) biomass available for use as animal feed as compensation for cool season grasses during the first half of July.     

Above-ground biomass estimation in ten tropical dry evergreen forest sites of peninsular India

The present study aims to estimate the above-ground biomass (AGB) distribution in ten 1-ha permanent plots, established in five sites each in inland and coastal tropical dry evergreen forests of peninsular India. Two linear regression equations, one using basal area (BA, Method 1) and the other using BA and height (Method 2) were followed. On using method 1, the AGB varied from 39.69 to 170.02 Mg ha⁻¹ and by method 2, it varied from 73.06 to 173.10 Mg ha⁻¹. The relationship between BA and AGB yielded a positive correlation for all the five sites of inland and coastal areas. The basic wood specific gravity of 41 tree species determined by oven-dry weight by volume, ranged from 0.46 to 0.92 g cm⁻³ for inland sites and 0.47 to 0.89 g cm⁻³ for the coastal sites. The AGB estimation obtained in this study represents a more realistic picture of biomass of tropical dry evergreen forests, because a relatively large area was sampled.     

Effects of wood ash and nitrogen fertilization on Scots pine crown biomass

The crown biomass, being one of the most susceptible components of the above-ground tree biomass, could respond positively to environmental changes and temporary increase in nutrient availability. The influence of wood ash and nitrogen fertilization on crown biomass was studied in a 40-year-old Scots pine (Pinus sylvestris L.) stand growing on a Haplic Arenosol. The 36-model trees for the crown biomass measurements were sampled for 3 growing seasons after the application of 5.0 t wood ash ha⁻¹, 180 kg N ha⁻¹, 2.5 t wood ash ha⁻¹ plus 180 kg N ha⁻¹, and control (untreated plots). The masses of the current, 1-year-old and older needles and shoots, and branches were measured. A significant influence on the current year needles and shoots was found after the application of 180 kg N ha⁻¹ or 2.5 t wood ash ha⁻¹ plus 180 kg N ha⁻¹. When wood ash was applied in combination with nitrogen, an extra response tendency of the crown growth, especially of the top and the middle sections of the crown, was determined. However, there was no short-term influence of wood ash on crown biomass growth.     

Effect of biomass species and plant size on cellulosic ethanol: A comparative process and economic analysis

The effects of five different biomass species and their chemical composition on the overall process efficiency and economic performance considering feedstock availability and feedstock costs to manufacture ethanol from lignocellulose were studied. First is a comparison of ethanol production and excess electricity generated between different biomass species. Results show that, at the same feedstock rate of 2000 Mg day^?1, aspen wood has larger ethanol production than switchgrass, hybrid poplar and corn stover, while the excess electricity generated is as follows in increasing order: aspen < corn stover < hybrid poplar/switchgrass. Second, our results show that the ethanol production is largely linear with holocellulose (cellulose plus hemicellulose) composition of the various biomass species. However, the relationship between excess electricity generated and non-holocellulose combustible component is nonlinear. Last, on environmental performance, it is found that the water losses per unit ethanol production are in the following order: aspen wood < corn stover < hybrid poplar < switchgrass. While corn stover is a potential feedstock to produce cellulosic ethanol with the lowest ethanol production cost at the present time, hybrid poplar and switchgrass are the two promising future energy crops.The effects of plant size analysis showed that the estimated feedstock delivered costs, ethanol production, excess electricity generated and solid and gaseous waste emissions all increase with plant size for the various biomass species. The ethanol production costs decrease with the increase in plant size with optimal plant sizes for corn stover in the range from 2000 dry Mg day^?1 to 4000 dry Mg day^?1.     

Effects of apical meristem loss on sylleptic branching and growth of hybrid poplar

The effects of apical meristem loss on the growth and development of hybrid poplar trees was investigated. This was done by clipping back either the apical meristem alone (dividing cells), or the apical meristem plus a small amount of additional stem tissue (expanding cells, <1 cm), at various times during the first growing season. Two clones (NM6–nonsylleptic habit, and DN34–slightly sylleptic habit) were tested at close spacing (0.6 m) in the nursery. Clipping generally increased the number of sylleptic branches formed. Clipping 69 days after planting resulted in the largest number of sylleptic branches while clipping 4 weeks later gave no increase in syllepsis. Clipping temporarily reduced height growth of both clones but total height at the end of the first growing season was not affected by any treatment. There were some slight differences in growth during the second growing season; despite these differences, total stem biomass and total tree biomass after 2 years were not affected by temporary loss of the apical meristem in the first growing season. Results suggest that death or removal of hybrid poplar apical meristems by tip borers or ungulates has no long-term effects on aboveground growth as measured by height or biomass.     

Economics of willow growing in Northern Ireland

This paper reviews the e economics of short rotation coppice willow as an energy crop in Northern Ireland. Gross margins are presented for willow production and compared with, in the particular circumstances of Northern Ireland, equivalent outputs from grain production, lowland sheep and suckler cow production. The model used indicated a gross margin of £45 ha⁻¹ yr⁻¹ for a 12 tDM ha⁻¹ annual coppice crop without subsidies where the crop value was placed at £40 t⁻¹. This was equivalent to a 7 t winter wheat crop at £70 t⁻¹ and compared favourably with both lowland sheep and suckler cows.Currently the industry in Northern Ireland is at a very early stage of development and this imposes cost penalties on the pioneer growers. This situation is compared with the situation in Sweden where there is an established industry of 15,000 ha, where costs are significantly lower. Gross margin for the pioneer grower in Northern Ireland is about £100 ha⁻¹ yr⁻¹ less than for Swedish willow growers.     

Experimental investigation of mixed convection heat transfer from longitudinal fins in a horizontal rectangular channel

Mixed convection heat transfer from longitudinal fins inside a horizontal channel has been investigated for a wide range of modified Rayleigh numbers and different fin heights and spacings. An experimental parametric study was made to investigate effects of fin spacing, fin height and magnitude of heat flux on mixed convection heat transfer from rectangular fin arrays heated from below in a horizontal channel. The optimum fin spacing to obtain maximum heat transfer has also been investigated. During the experiments constant heat flux boundary condition was realized and air was used as the working fluid. The velocity of fluid entering channel was kept nearly constant (0.15 ? w_in ? 0.16 m/s) using a flow rate control valve so that Reynolds number was always about Re = 1500. Experiments were conducted for modified Rayleigh numbers 3 × 10⁷ < Ra¹ < 8 × 10⁸ and Richardson number 0.4 < Ri < 5. Dimensionless fin spacing was varied from S/H = 0.04 to S/H = 0.018 and fin height was varied from H_f/H = 0.25 to H_f/H = 0.80. For mixed convection heat transfer, the results obtained from experimental study show that the optimum fin spacing which yields the maximum heat transfer is S = 8–9 mm and optimum fin spacing depends on the value of Ra¹.