Index | Search | Home | Table of Contents

link to pdf file

Lewis, M. and M. Jackson. 2002. Nalgrass: A nonwood fiber source suitable for existing US pulp mills. p. 371–376. In: J. Janick and A. Whipkey (eds.), Trends in new crops and new uses. ASHS Press, Alexandria, VA.

Nalgrass: A Nonwood Fiber Source Suitable for Existing US Pulp Mills

Mark Lewis and Michael Jackson


Arundo donax, Poaceae, commonly known as “nalgrass,” is a member of the reed family that grows both wild and cultivated in California. Evaluated as a potential paper fiber source in Italy and Hungary over 50 years ago and included in the 1970s USDA evaluation of potential annual fiber crops, the high potential of A. donax has not been widely recognized. A new evaluation was begun in 1999 by Nile Fiber Pulp & Paper Company. The pulping behavior and pulp properties of the material were tested. The results were surprising in that many of the attributes of A. donax were found to be suitable for direct substitution for hardwoods in existing kraft mills without major equipment changes. This is in contrast to most other nonwood raw materials, which required specialized systems for handling and digestion in order to be converted to bleached pulp. This report presents data from kraft pulping of nalgrass and shows how its behavior closely parallels that of wood.


Arundo donax is a reed species native to southern Europe. It was introduced into North America in the late 1800s and now grows wild in California and other southern regions. Its growth area in the US is the southern states and it is considered a weed species in California but is grown as a cultivated crop for the production of musical instrument reeds. Literature on the pulping of nalgrass is limited. It is a reed species and other reed types are utilized for papermaking, particularly in China. Limited information is available on these other reed species, however most references are in obscure and unavailable foreign literature.

Considerable effort was made in the 1940s and 1950s in Hungary and Italy to cultivate and utilize material from reeds. Nalgrass was selected as the reed of choice in Italy. However no commercial operations resulted from this effort with nalgrass, although in Hungary some native reed material was utilized in pulp and paper manufacture.

In the program beginning in the 1970s to identify alternative commercial crops, the USDA screened a wide variety of species as possible crops for fiber production. Nalgrass was included but testing on it was not advanced beyond the collection of basic data. This program ultimately resulted in the selection of kenaf as a desirable crop species.

The common reed, Phragmites communis, and other related reeds are utilized worldwide in small quantities in papermaking; in particular, China has a significant utilization of reed in papermaking operations. Various pulping processes have been used or tested for the pulping of common reeds for papermaking fibers and for conversion into viscose rayon. The conclusion drawn by Wiedermann (1920) was that the reeds offered a good opportunity to make acceptable papermaking fiber, but that serious problems were identified in cultivation and harvesting with mechanical equipment. Pulping yields were at acceptable levels, above those typical of other nonwood fibers, and brightness was high in unbleached pulp, resulting in easy bleaching to high brightness levels. Pulp strength properties were similar to hardwood kraft species. The relatively high silica content typical in reeds and other nonwood species was recognized as a problem in waste liquor recovery, although means of alleviating these problems exist.


Material for the present studies was cut from both fresh and aged growths in Orange County, California. The nalgrass stem has a dense ring of tissue surrounding a hollow core. Stem diameters are typically 3/4 to 1-3/4 inches in diameter. It can be cut or milled into lengths similar to wood chips and once crushed to break the circular cross section has bulk density similar to that of wood chips (Table 1).

Table 1. Bulk density of nalgrass, wheat straw, and northwest softwood.

Material Bulk density (lb/ft3)
Uncompacted green Compacted green
Nalgrass 10.8 12.5
Wheat straw 2-6 3-7
Northwest softwood 12-14 12-15

For the kraft pulping trials tests were made at four different cut lengths, 1/2, 3/4, 7/8, and 1-1/4 inch. A sample was also prepared by cutting the nalgrass into flakes in a slicer used for strand board particle preparation. A sample of dried material was also included. This was cut to 7/8 inch length and was included to evaluate whether liquor penetration was hindered by drying as is the case with wood chips.


Kraft Pulping

Kraft cooking and soda pulping of the nalgrass material was completed using a pilot digester system. Cooks were made with each of the chip samples under conditions aimed at producing delignification to the 20 Kappa level suitable for bleaching. Pulping conditions are given in Table 2. Subsequently, an extensive number of cooks have been made to produce larger quantities of pulp for papermaking evaluations.

Table 2. Kraft pulping conditions for chips size and type evaluation (chips as is and presteamed 100 psig, 5 min).

Chip condition Chip
EA (%) Sulfidity
Kappa no. Rejects
Chips as is
Fresh stem 1/2 850 170 4.5:1 15 2.4 17.4 1.1 38.3 10.3
Fresh stem 3/4 850 170 4.5:1 15 2.4 41.0 0.9 32.8 10.0
Fresh stem 7/8 850 170 4.5:1 15 2.4 17.6 3.6 46.0 8.5
Fresh stem 1 1/4 850 170 4.5:1 15 2.4 18.2 3.2 37.3 8.1
Dry stem Flaked 850 170 4.5:1 15 2.4 14.6 0.2 53.4 7.1
Dry stem 7/8 850 170 4.5:1 15 2.4 14.9 3.3 46.2 11.2
Chips presteamed
Fresh stem 7/8 850 170 4.5:1 15 2.4 14.2 2.2 43.3 -
Dry stem 7/8 850 170 4.5:1 15 2.4 14.2 2.5 44.1 -
Soda pulping
Fresh stem 7/8 850 170 4.5:1 15 0 16.0 2.1 46.3 12.1


The bleaching of kraft pulps was made with Elementary Chlorine Free (ECF) sequences. Bleaching tests were made on pulp from a larger scale cook on the 7/8 inch cut material using an ECF bleach consisting of Chlorine dioxide (Do). Extraction with oxygen and peroxide (Eop). Chlorine dioxide (D1). The results are shown in Table 4.

Table 3. Pulping conditions for 7/8 inch chip size compared to wood chips.

Material H-Factor Temp (C) Liquor/material EA (%) Sulfidity (%) Kappa no. Rejects (%)
Nalgrass 7/8"
Hardwood typical
Softwood typical

Table 4. Bleach response—kraft pulp, D.Eop.D bleach sequence—0.20 Kappa factor, soda pulp D.Eop.D bleach sequence—0.25 Kappa factor.

Stage Consistency
Time (min) Temp (C) O2 (psi) H2O2 (%) NaOH (%) ClO2 (%) pH Brightness
Kraft pulp, D.Eop.D bleach sequence-0.20 Kappa factor
Do 10 0.2 30 60 -- -- -- 1.34 3.3 --
Eop 10 -- 90 100 30 0.7 1.7 -- 9.5 --
D1 10 -- 120 70 -- -- -- 1.5 3.4 83.94
Kraft pulp, D.Eop.D bleach sequence-0.25 Kappa factor
Do 10 0.25 30 60 -- -- -- 1.68 3.3 --
Eop 10 -- 90 100 30 0.7 1.7 -- 9.5 --
D1 run 1 10 -- 120 70 -- -- -- 1.25 3.4 85.6
D1 run 2 10   120 70       1.5   86.4
Soda pulp D.Eop.D bleach sequence-0.25 Kappa factor
Do 10 0.25 30 60 -- -- -- 1.47 3.2 --
Eop 10 -- 90 100 30 0.7 1.7 -- 9.8 --
D1 10 -- 120 70 -- -- -- 1.25 3.3 82.9

Initially, a chlorine dioxide charge in the first stage of 0.20 Kappa factor (percentage equivalent chlorine/Kappa number) was applied, followed by 1.5% chlorine dioxide in the third stage. This resulted in a brightness of 83.8, Table 4. Modification to a 0.25 Kappa factor application in the first stage resulted in brightnesses of 85.6 and 86.4 with 1.25% and 1.5% chlorine dioxide in the third stage, respectively, Table 4.

A total chlorine dioxide charge of 3.18% was required for the 86.4 brightness. In the Econotech tests a brightness of 90.0 was reached in a five stage bleach using 4.34% chlorine dioxide. Softwood kraft pulps typically require 5.8–6.2% chlorine dioxide to reach a brightness level of 90.0.

The bleach response of the soda cook pulp is shown in Table 4. A brightness of 82.9 was reached with a total chlorine dioxide charge of 2.72%. Indicating that the soda pulp responds to bleaching similar to the kraft pulp.

Handsheet Properties

Standard testing of pulp properties was made using TAPPI procedures. Pulp from the 7/8-inch chip sample was beaten in a PFI mill to various freeness levels. Handsheets were made from the refined 7/8 inch cut nalgrass pulp at several freeness levels and tested for strength properties, (Table 5). Pulps from the other chip cut lengths were beaten to the 400 ml CSF level for comparison.

Table 5. Handsheet strength tests.

Chip size
PFI (k) Freeness
Burst index
Tear index
Tensile index
Kraft pulp            
7/8 0 700 2.5 4.0 41.6  
  1 605 3.8 10.4 63.5  
  2 488 4.8 9.4 72.4  
  3 415 5.1 9.2 78.9  
  3.2 404 4.5 9.4 75.1  
  3.6 391 5.0 9.4 78.3  
1/2 0 733 2.6 4.7 39.4  
  3.2 413 4.8 8.8 77.6  
3/4 0 700 3.1 4.0 49.8  
  3.2 393 5.1 9.3 80.0  
1 1/4 0 709 3.1 4.2 47.2  
  3.2 393 5.3 9.0 81.2  
Soda pulp            
7/8 0 720 2.1 11.0 43.3 1.74
  1.6 500 3.9 11.7 53.0 1.47
  3.2 380 5.1 10.9 68.3 1.41


The character of chipped material is critical to the processing behavior in conventional pulping equipment. The bulk density of the chipped material is important in terms of packing into digesters and sizing of conveyors and other process equipment. In the digester bulk density dictates the digester capacity, available cooking time, and the potential liquor-to-wood ratio. The high bulk density of chipped nalgrass will allow it to be processed in conventional, existing chip handling and pulping equipment. Cooking liquor to raw material ratios can be low, similar to those used for wood chips, resulting in high waste liquor concentrations.

Another important chip characteristic is the ability of the cooking chemicals to penetrate into the center of the chip during pulping. The initial tests were made by hammermilling the reed to prepared chips. After screening to remove fines and oversized material the accepts were used for pulping. The material gave pulp with low uncooked rejects, indicating that the penetration of cooking liquor was uniform. In the present study different chip length material was pulped to evaluate possible interaction of different chipping methods with the pulping process.


All samples cooked with similar results. The cooking time is short as indicated by the low H Factor (a chemical reaction value combining temperature and reaction time). Cooking times were up to half those of softwoods. The high bulk density of the nalgrass chips also allowed use of a low liquor-to-chip ratio similar to that used for wood chips. This indicates that nalgrass pulping could be made in the same equipment as wood chips and with the same heat economy. Typical low-density straw and other nonwood plant material require high liquor to wood ratios although cooking is rapid as found with this nalgrass material.

The four various lengths of chips show only small, and probably insignificant, differences in pulping response. Although the 3/4 inch chip had slightly lower Kappa, 14.0 vs. 17.6–18.2 for the longer chips, the 1/2 inch chip gave 17.4 Kappa. The uncooked rejects were lower in the short cut chips, 0.9%–1.1%, compared to the longer chips, 3.2%–3.6%, but these levels are low, indicating that uniform penetration of cooking liquors into the material occurred and also showing that the nodes cooked well. The nodes of grasses, of which nalgrass is a member, are sometimes resistant to pulping.

The flaked chips cooked similarly to the saw cut chips, giving a low Kappa, 14.6, and low rejects, 0.2% (Table 2). This type of chip preparation would be satisfactory for commercial operations.

The dried material showed a pulping response similar to the fresh material, Kappa 14.9, rejects 3.3% (Table 2), indicating that there are no problems with the penetration of liquor into dry nalgrass chips. In addition, presteaming the chips did not significantly change the pulping response, (Table 2), indicating that liquor penetration occurred rapidly and completely in both the green and dry materials. These results show that chips could be used from fresh or dry material without significant changes in process conditions.

The pulping of the 7/8 inch cut nalgrass is compared to typical hardwood and softwood kraft pulping in Table 3. The nalgrass cooks more rapidly than both types of wood, requires less chemical and produces only slightly higher rejects (not a significant difference).

As seen in the kraft cooks described above, the nalgrass material cooks easily in the kraft process. This is typical of nonwood materials, and in fact, many mills using nonwoods use the soda cooking process rather than the kraft. The soda process uses caustic soda without the addition of sulfur to form sulfide, the key to the kraft process. Softwoods do not pulp readily by the soda process and require the added sulfide of the kraft system. It is advantageous to be able to use the soda process as the foul smelling sulfur compounds are avoided and the recovery process simplified. Tests were made using the soda process to pulp the nalgrass material.

The results shown in Table 2 are essentially equivalent to the pulping with the kraft process. A Kappa of 16.0 was obtained with the same H Factor as in the kraft tests and with Effective Alkali charge slightly increased, from 15% to 17%. Rejects were in the same range as in the kraft cooks and the yield was high, 46.3%.

Paper Testing

The initial pulp freeness before beating was 700 ml CSF which is a very high and desirable level compared with typical nonwood material. This compares to >700 ml for softwood pulps and 600–650 for hardwood pulps and are advantageous, allowing the papermaker to modify the pulp properties without restriction and to allow high drainage in the papermaking operation.

The handsheet strength measurement, burst, tensile, and tear, from the current series of tests are all at favorable levels and higher than those obtained in the Econotech tests. Comparison of the two sets of results from nalgrass and from typical wheat straw, kenaf, hardwood, and softwood are shown in Table 6. The nalgrass has remarkably high strength in all categories. The sheet bulk is high compared to hardwoods which indicates the material has significantly different characteristics than the straws.

Table 6. Comparison of nalgrass with other pulps.

Material Freeness
PFI mill
Burst index
Tear index
Tensile index
Nalgrass UW 400 3200 4.5 9.4 75 1.40 86
Nalgrass Econotech 400 900   8.7 53 1.59 90
Wheat straw 400 400 -- 3.7 40 1.24 85
Whole kenaf 400 -- 5.5 10 65 -- --
Aspen kraft 400 464 2.1 7.6 46 1.43 89
D fir kraft 400 8100 6.8 22.4 92 1.81 89

Handsheet tests from the bleached soda pulp are shown in Table 5. The results show slightly lower burst and tensile but higher tear strength. This is typical of soda pulps compared to kraft.


Material from Arundo donax, termed “nalgrass” was prepared for pulping by hand and processed on a pilot scale by the kraft, soda, and alkaline peroxide mechanical pulping processes, APMP. The kraft and soda pulps were laboratory bleached by a three stage ECF sequence (D.Eop.D). The pulps produced were evaluated for papermaking properties by the preparation of handsheets after beating in a PFI mill and were tested for standard strength and optical properties.

Based on the results obtained and previous evaluations made in the Econotech Services laboratory the following conclusions can be made:

  1. The Nalgrass material cooks rapidly in the kraft process to low Kappa levels with yield in the range of 33%–53%.
  2. Chip length, raw material moisture content and use of presteaming had very little effect on the response in kraft cooking, indicating that material preparation is not critical.
  3. Flaked material from strand board slicing also pulped readily.
  4. Bleach response for the kraft pulp was good requiring less chemical in an ECF, D.Eop.D, bleach sequence to reach a given brightness than typical softwood. Brightness of 86.4 was reached in the three stage bleach sequence.
  5. Strength properties of handsheets from the bleach material were higher than for typical hardwoods and significantly above typical nonwood materials.
  6. The material pulped well under soda cooking conditions showing comparable pulping and bleaching results to the kraft system with strength slightly modified to lower tensile and higher tear.
  7. Alkaline peroxide mechanical pulping proceed well with the nalgrass but the handsheet test data was very poor. It was concluded that the wax content of nalgrass caused interference with fiber bonding in the hand sheets. Further investigation is required to resolve this issue.


Based on the kraft and soda pulping results, it is recommended that mill scale trials be made and paper produced with the nalgrass material substituting for hardwood and possibly some of the softwood content of the furnish. Further investigation of the APMP response should be made as this process has high promise of a low capital, low effluent add-on system for existing mills. The nalgrass material can be expected to give good results in most pulping processes and its response in semi-chemical processes for corrugating medium and linerboard should be demonstrated.