Art$i'ciu/ Orycrns 19(11): I 1 7 6 1 180, Blackwell Science. Inc., Boston 0 1995 International Society for Artificial Organs

The Influence of the Dialysate Flow Rate on Hollow Fiber Hemodialyzer Performance

Richard Allen, Thomas Herbert Frost, and *Nicholas Andrew Hoenich

Department of Mechanical Materials and Manufacturing Engineering, and *Department of Medicine, University of Newcastle upon Tyne, Newcastle upon Tyne, U . K .

Abstract: The clearance characteristics of two sizes of hemodialyzers (0.9 m2 and 1.5 m') from the same range of products have been studied over the dialysate flow range of 50O-3,OOO ml/min to establish the device's overall mass transfer resistance characteristics. The results obtained demonstrate a difference in the overall mass transfer re- sistance which is most marked at the commonly used dialysate flow rate of 500 mlimin. This difference suggests that the increase in size results in the introduction of flow imperfections which reduces the benefit that might be gained by the use of a larger surface area. Results estab- lished indicate a reduction in the overall mass transfer resistance with an increasing dialysate flow rate. This re-

duction is attributed to the presence of turbulence in the dialysate pathway at higher flow rates. The presence of such turbulence was confirmed by visual inspection of the dialyzer after the completion of the study when it was noted that the original well-ordered configuration present in a new dialyzer had been substantially disturbed. Cor- relation of the dialysate flow rate with overall mass trans- fer resistance by the use of a Wilson plots indicates a nonlinear relationship. This nonlinearity is attributed to a nonfully developed turbulent flow profile in the dialysate pathway. Key Words: Hemodialyzer-Mass transfer resistance.

Hollow fiber hemodialyzers now account for over 90% of the clinical treatments for chronic renal fail- ure. Their popularity is attributed to their high per- formance, low and stable blood compartment vol- ume, and ease of set up and use.

Hollow fiber hemodialyzers consist of a bundle of fibers in a perspex tube and have a construction similar to a shell and tube heat exchanger. Blood flows, via headers throlugh the lumen of the fibers, while dialysis fluid flows on the outer side of the fibers in a countercurrent mode, entering and leav- ing the device via ports incorporated into the per- spex outer casing. An analysis of the performance of two dialyzers of the same type, but differing in their surface area, have been undertaken in the lab- oratory to assess their performance with respect to urea clearance over 50O-3,OOO ml/min of dialysate flow rate range. The data obtained were used to

Received May 1995. Address correspondence imd reprint requests to Dr. N. A.

Hoenich, Department of Medicine, Medical School, University of Newcastle upon Tyne, Frarnlington Place, Newcastle upon Tyne NE 2 4HH, U.K.

determine the overall mass transfer resistance char- acteristics.

The purpose of the study was to investigate if the increase of surface area is associated with an in- crease in overall mass transfer resistance resulting from flow imperfections in the fiber bundle.

MATERIALS AND METHODS An in vitro test system was constructed for the

purposes of the study. The flow rates in the blood pathway were maintained by the use of a Gambro BPlO blood pump while in the dialysis circuit, an AC Bronze Mini-Puppy pump (ITT Jabsco, Hoddeston, U.K.) was used in conjunction with a valve controlled rotameter to generate and control the flow in the dialysate pathway. The flow rates during the experiments were determined by volu- metric timed collections. Pressures in both circuits were monitored at the inlet and outlet to the dia- lyzer by the use of mercury manometers and cor- rected to the center of the dialyzer. In-line heaters were used to maintain the temperature at 37°C in both circuits.

The dialyzers studied were the Gambro GFE 9

1176

DIAL YSATE FLOW RATE AND HEMODIAL YZER PERFORMANCE 1177

and GFE 15 hollow fiber devices with surface areas of 0.9 and 1.5 m2, respectively. The increase in sur- face area was achieved by using more and longer fibers. Both types of dialyzers used an identical membrane (Cuprophan: 11 pm wall thickness, 200 p,m internal diameter).

The clearance characteristics of the dialyzers as a function of both blood and dialysate flow rate were established on two dialyzers of each type. A sepa- rate series of measurements to establish the dialyz- ers' ultrafiltration characteristics were performed and used to correct the clearance to zero ultrafiltra- tion. Samples taken from the blood and dialysate circuits were analyzed using a Beckmann Astra multichannel analyzer, and the results were used in the calculation of clearance by the conventional for- mula. Mass balance errors were less than 5% for the studies.

RESULTS AND DISCUSSION

The variation of clearance (KB) with both blood (QB) and dialysate flow (QD) rates is shown in Fig. 1. For a dialyzer using a contraparallel flow config- uration, in which the two flow rates are unequal, the dialyzers' effectiveness may be defined as

QB

QD 1 - - e x p - Z QB

where

Surface area ( S ) QB Overall mass transfer resistance (RT) Z =

which when rearranged gives

L' -

The overall mass transfer resistance in a dialyzer is made up of additive contributions from the blood and dialysate boundary layers and the membrane. The reciprocal of the overall mass transfer resis- tance is the overall mass transfer coefficient (ko). The mass transfer coefficient area product (koA) is a parameter that has been used to determine the effects of increasing blood and dialysate flows on

device performance. In using this concept, it is gen- erally assumed that koA remains constant over the range of blood and dialysate flow rates of interest. The variation of overall mass transfer resistance at a blood flow rate of 300 ml/min with a dialysate flow rate for the two sizes of dialyzer studied are shown in Fig. 2.

For hollow fiber dialyzers, the resistance in the blood pathway may be calculated by a technique described by Sigdell (1). This technique was applied to the dialyzers studied, and the contribution due to the blood layer established (5 minlcm). The mem- brane resistance for Cuprophan was taken from published literature and adjusted for wet membrane thickness (1,2). Using these values the relative con- tribution of the dialysate pathway resistance to the overall value was calculated. These results indicate that the contribution from the dialysate pathway di- minishes with an increasing dialysate flow rate.

To obtain a more quantitative estimate of the con- tribution of the individual phases to the overall mass transfer resistance, the overall mass transfer resistance was plotted as a function of the dialysate flow by the constructing a Wilson plot ( 3 ) . The di- alysate flow was assumed to be turbulent, in which case R, is proportional to (Q,)po.8 (4). For the GFE 9, a linear relationship was observed, for the GFE 15 the relationship was curvilinear (Fig. 3 ) . Since the theory behind this method involves very few assumptions for rigid turbulent systems, the devia- tion from linearity may be a result in a lower degree of turbulence than expected. At infinite dialysate flow rates, the contribution from the dialysate path- way approaches zero, and the intercept represents the sum of the contributions from the membrane and the blood pathway. The negative values noted at high dialysate flow rates are probably a conse- quence of the errors in the measurement of clear- ance.

CONCLUSIONS

A number of points emerge from this experimen- tal study. First, there is a considerable difference in the overall mass transfer resistance between two similarly constructed hemodialyzers with different surface areas. The consequence of this difference is that the improvement in performance due to the increase in surface area is lower than theoretically possible. The difference for the dialyzer types stud- ied is most marked at the commonly used dialysate flow rate of 500 ml/min. The presence of such flow imperfections in multiple layer parallel plate de-

Artif Organs, Vol. 19, No. 11, 1995

1178 R. ALLEN ET AL.

400

.E 300 E 1. E

Q 200 - 0 c m m Q

L

- 0 100

Garnbro GFE9 Haemodialyser

0 Blood Flow 200 mlfmm Blood Flow 300 mlfrnin

A Blood Flow 400 rnl/min

0 1 0 1000 2000 3000

Dialysate Flow (rnl/rnin)

Gambro GFE15 Haemodialyser

400 1

0) 200 0 c m m Q

L

- 0 100 0 Blood Flow 200 ml/rnln

Blood Flow 300 mlfmin A Blood Flow 400 mlfmln

0 0 1000 2000 3000

Dialysate Flow [ml/min)

vices has been previously described (3, but this is the first description in hollow fiber devices.

Second, the overall mass transfer resistance of the dialyzers decreases with increasing dialysate flow rate. This reduction is attributed to the pres- ence of turbulence in the dialysate pathway at higher flow rates. The presence of such turbulence was confirmed by visual inspection of the dialyzer after the completion of the study when it was noted that the original well-ordered configuration present

FIG. 1. Variation of urea clearance is shown over the dialysate flow rate 500-3,000 mlhnin for the two sizes of hemodialyzers studied.

in a new dialyzer had been substantially disturbed. The Wilson plots, however, indicate that the degree of turbulence varies between the dialyzers and was absent in the smaller size of the devices studied.

Third, increase of the dialysate flow rate results in an increased performance; however, this in- crease, as demonstrated by the work of Sigdell and Tersteegen, is maximal when used in conjunction with high blood flow rates (6). The use of dialysate flow rates in excess of 1,000 ml/min is expensive

A r t f O r g a n s . Vo l . 19. No. 11, 1195

DIAL YSATE FLOW RATE A N D HEMODIALYZER PERFORMANCE 1179

\ c E

a ...........

l-

l o 1 RM + RB

..............

Blood Flow 300 ml/min

I 1

0 1000 2000 3000 0

Dialysate Flow (rnl/minl

4o , Gambro GFE15

FIG. 2. Variation of the overall mass transfer resistance is shown over the dialysate flow rate 50C-3,OOO rnlirnin for the two sizes of hernodialyzers studied.

10 4 Blood Flow 300 ml/rnin

0 1 I I I

0 1000 2000 3000

Dialysate Flow (ml/minl

Acknowledgments: This work formed part of the final year project dissertation submitted by R. Allen in partial

Mechanical Engineering, the University of Newcastle

in clinical practice. An alternate approach might be

enhance solute In such a system the con- to use a recirculating sing1e pass dialysate system to fulfillment of the regulations for Honours Degree of

tents O f a Small reservoir, Similar to the Volume con- tained in the dialysate compartment of the dialyzer, are recirculated through the dialyzer at a high rate, and fresh dialysis fluid is added to the reservoir at a constant rate which is identical to the amount of dialysate flowing to waste after recirculation.

upon Tyne. The help OF Miss Celia Woffindin, Dialysis Research Laboratory, Department of Medicine Univer- sity of Newcastle and the staff of the Department of Clin- ical Biochemistry Royal Victoria Infirmary and Associ- ated NHS Trust, Newcastle upon Tyne are gratefully acknowledged, together with the help of ITT-Jabco who provided the pump for the studies.

ArtifOrganA, Vol. 19, No. I I . 1995

1180 R . ALLEN ET AL.

Wilson Plot Garnbro GFE9

30 1 25

20 - E

E

0 \ c

15 .- I

a' 10

I I I I

0.002 0.004 0.006 0.008 0 1

0

30

25

20

15

10

5

0

FIG. 3. Wilson Plots for the two sizes of hemodialyzers studied are shown.

Garnbro GFE15

.

I I I I 0 0.002 0.004 0.006 0.008

REFERENCES

1. Sigdell JE. Operating characteristics of hollow fibre dialy- sers. In: Nissenson AR, Fine RN, Gentile DE, eds. Clinical dialysis. Connecticut Appleton Lange 1990:97-111.

2. Colton CK. A review of the development andperformance of hemodialysers. Artificial Kidney-Chronic Uremia Program, National Institute of Arthritis and Metabolic Diseases. Fed- eral Clearinghouse Accession No. PB 183-281. US Depart- ment of Health Education and Welfare, Washington, 1967.

3. Wilson EE. A basis for the rational design of heat transfer apparatus. Trans Am Soc Mech Eng 1915;37:47.

4. Hoenich NA, Frost TH. Influence of design and operating variables on conventional haemodialysis. In: Whelpton D, ed. Renal dialysis. London, Sector Publishing, 1974:85-104.

5 . Hoenich NA, Conceicao S, White T, Ward MK, Kerr DNS. Large surface area dialysers-A question of performance? Proceedings of the European Society of Artificial Organs 1976;3 : 185-90.

6. Sigdell JE, Tersteegen B. Clearance for a dialyser under varying operating conditions. Artif Organs 1986;10:219-25.

ArtifOrgans, Vol. 19, No. 11. 1995