ML20235N083
| ML20235N083 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 02/22/1989 |
| From: | Gridley R TENNESSEE VALLEY AUTHORITY |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 8903010025 | |
| Download: ML20235N083 (22) | |
Text
,.
TENNESSEE VALLEY AUTHORITY CHATTANOOGA, TENNESSEE 37401 SN 157B Lookout Place FEB 221989 U.S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, D.C.
20555 Gentlemen:
In the Matter of
)
Docket Nos. 50-327 Tennessee Valley Authority
)
50-328 SEQUOYAH NUCLEAR PLANT (SQN) - EVALUATION OF SAFETY PARAMETER DISPLAY SYSTEM (SPDS)
Enclosed is TVA's response to a verbal request from H. Garg, NRC, of January 13, 1989, for a more detailed status of TVA's actions to resolve the audit findings on SPDS.
If you have any q0estions concerning this submittal, please telephone M. A. Cooper at (615) 843-6549.
Very truly yours, TENNESSEE V LEY AUTHORITY R. Gridley, anager Nuclear Licensing and Regulatory Affairs Enclosure cc: See page 2 I
8903010025 890'222 PDR ADOCK 05000327 r
p PDR Ij An Equal opportunity Employer
-I,
U.S. Nuclear. Regulatory Commission FEB 221888 cc (Enclosure):
Ms. S. C. Black, Assistant-Director forProjects TVA Projects Division' U.S. Nuclear Regulatory Commission One White-Flint, North 11555 Rockville Pike Rockville, Maryland-20852 Mr. F. R.'McCoy, Assistant Director for Inspection Programs TVA Projects Division x
U.S. Nuclear Regulatory Commission Region II i
101 Marietta Street, NH, Suite 2900 L
Atlanta,: Georgia 30323 Sequoyah Resident Inspector Sequoyah Nuclear Plant 2600 Igou Ferry Road Soddy Daisy, Tennessee 37379 g
I LL__--
l Enclosure l
"3.1 Concise display of critical plant variables to control room operators Neither the 6 CSF displays nor any of the other displays under the SPDS i
map contain all of the CSFs required by NUREG-0737, Supplement 1.
The CSFs do not address the function of Radioactivity Control. Also, the 6 CSF displays do not use inputs from containment Hydrogen Concentration Reactor Vessel Level Indication System or Main Steam Line Monitors.
The review team evaluated the OTSC displays to determine if one of them might contain all of the NUREG-0737, Supplement i safety functions or parameter inputs used to monitor the safety function. The 2 PSI (Plant Status) display under the OTSC map came very close in terms of parameter input, but still lacked some parameters (containment sump level and hydrogen concentration) and does not display the CSFs.
In addition to the fact that no single display could be identified which adequately includes all safety functions, it was noted that the algorithms which drive the CSF displays do not run continuously when the plant is operating at power. The CSF algorithms only start operating about one minute after a reactor trip."
TVA Response SQN has completed a significant design change in the SPDS since the NRC audit in March 1988. The status for NUREG-0737, Supplement 1, critical safety functions (CSF) is now displayed on SPDS and onsite Technical Support Center (OTSC) screens in the main control room (MCR) and 0TSC at all times (i.e., not just posttrip). This is accomplished with an array of status blocks located in the bottom left corner of every SPDS and OTSC screen (refer to figures 1 Each of the ei ht blocks corresponds to an SPDS status tree. A through 8).
3 new radioactivity control status tree has been added, and containment hydrogen concentration has been added to an existing tree. Addition of reactor vessel level will ba complete when reactor vessel level indicating system (RVLIS) versions of the emergency operating procedures (E0Ps) are implemented before the restart of unit 2 from the ongoing third refueling outage.
The changes to the SPDS that are required for the RVLIS E0P changes are described below.
Main steam line radiation monitors will be added to the radioactivity control status tree when the main steam line radiation monitors are added as part of the postaccident monitoring (PAM) upgrade scheduled for the cycle 4 refueling outages.
Current Design There are eight critical safety function status trees (see figures 1 through 8); they are described below with regard to how they have been modified since the audit.
In addition to the changes noted below for each status tree, three generic changes have been made:
1.
The CSF SAT label on satisfied terminal blocks has been relabeled OK to match the hard copy status trees.
2.
A 4 by 2 array of blocks indicating the status of the eight status trees is added to the bottom of.the displays. This makes these SPDS status tree displays consistent with the other SPDS and OTSC displays.
3.
Two new blocks for the new radioactivity and post-LOCA (loss of coolant accident) decay heat removal have been added to the left side of each SPDS status tree display.
Suberiticality: This is similar to the suberiticality status tree presented during the audit, but an entry point decision block has been added to permit the use of the tree while operating at power. The decision block checks if the reactor trip breakers are closed.
If they are closed, then the assumption is the reactor plant is configured for power operation; therefore, subcriti;ality is not a safety requirement.
Core Cooling: This is the same as the core cooling status tree presented during the audit with the exception of changing the setpoint for reactor coolant system (RCS) subcooling margin. The subcooling margin block is now identified as adequate if the reactor is tripped and the margin is at least 40 degrees Fahrenheit (F) or at least 15 degrees F if the reactor is not tripped. The position of the reactor trip breakers is used to determine if the reactor is tripped. This status tree is being modified for the RVLIS version of the SPDS as noted below.
Heat Sink: This is the same as the heat sink status tree presented during the audit.
Pressurized Thermal Shock: This is the same as the pressurized thermal shock status tree presented during the audit with the exception that the temperature setpoints have been corrected to match the hard copy status trees.
Containment: This is similar to the containment status tree presented during the audit. A new status block for hydrogen concentration being less than 0.5 percent has been added to the tree.
The setpoints on the containment sump and radiation levels have also been modified to agree with the setpoints in the hard copy status trees.
Inventory:
This is the same as the inventory status tree presented during the audit.
It is being modified for the RVLIS SPDS version as noted below.
Effluent and Area Radioactivity:
This is a new status tree that provides indication if significant radioactive. leases to the environment are occurring.
It monitors containment radiation, auxiliary building radiation and effluent, shield building effluent, service building effluent, and secondary plant radiation levels.
- __ _- _ _- A
Decay Heat Removal:
This is a new status tree that checks for removal of decay heat from the containment after a LOCA.
It checks the position of the containment sump recirculation valves; if both are shut, the containment is assumed to be in a condition not requiring post-LOCA heat removal.
If either one is open, residual heat removal (RHR) flow is checked to ensure the containment sump water is being recirculated through the RHR system. NRC (C. Goodman, R. Eckenrode, and F. Orr) concurred with this approach on September 13, 1988, during a telephone conversation with TVA (R. Quirk and R. King).
RVLIS Design Two status trees are being modified for the RVLIS design:
core cooling and inventory. The changes are described below.
Core Cooling:
Four decision blocks are added to incorporate RVLIS in the core cooling SPDS status tree. After ensuring that the maximum core exit temperature is less than 1,200 degrees F and there is adequate RCS subcooling, the status is called OK.
If there is inadequate subcooling, a check is made to determine if any reactor coolant pumps (RCPs) are running. Depending on the result of this determination as well as the maximum core exit temperature and the reactor vessel level, the operator is directed to appropriate emergency procedures.
Inventory: The check for pressurizer level being normal is changed to a check that the reactor vessel level is adequate.
System Operation Modifications The operation of the SPDS function key has been modified so that it is now used to request the display of the SPDS status tree in the highest state of a1.a rm.
If no SPDS status trees are in alarm (i.e., the eight status trees are in an OK status), the subcriticality status tree is displayed. A bar chart function key has been added to request the display of the bar charts that were formerly displayed when the SPDS function key was depressed.
When an SPDS status tree goes into alarm, the associated status block on every SPDS and 0TSC screen starts to flash.
An alarm acknowledge key has been added to permit these alarms to be acknowledged. This alarm acknowledge feature is inhibited on keyboards outside of the MCR.
I "3.3 Continuous display of plant safety status information l
Currently, there is no requirement or procedure to continuously display any specific SPDS screen.
In addition, the s-reens as they now exist do not proviGe valid CSF information until 60 seconds after the reactor trip breakers are open.
Further, none of the individual screens provides information about all 5 CSFs."
I 1
-d
l TVA Response As noted in the response to item 3.1, the status of each CSF is now continuously displayed on the SPDS and OTSC screens. Therefore, there is no need to specify which screens must be displayed at any time. Also, as noted l
in the response to item 3.1, valid CSF information is now provided both pretrip and posttrip.
"3.4 Should have a high degree of reliability 3.4.1 Data Validity All SPDS/0TSC computer points undergo a basic instrument range check.
Most composed (computed) points undergo more sophisticated validation in the form of redundant sensor algorithms.
One composed point which should be considered for further validation is Subcooling Margin.
The Subcooling Margin algorithm uses the auctioneered hottest Core Exit Thermocouple (CET) to compare against a computed saturation temperature.
Because 4 to 5 CETs in the core of each unit are failed (some high, some low), the failed or spuriously high CETs are used in computing Subcooling Margin. This problem could be eliminated by using an averaged, validated high CET computer point."
TVA Response The worst-case thermocouple is used because there is a significant difference between temperatures near the conter and periphery of the core. The use of the worst-case thermocouple--actual, failed, or spuriously high--is conservative and, thcrefore, is still part of the SQN SPDS design.
In the case where a thermocouple fails (high or low or some other value), the individual thermocouple can be taken out of scan; and a manual value can be substituted. This can be accomplished from any cathode-ray tube (CRT) on the system in approximately one minute.
Once the point is taken out of scan, it ir not used in the highest core exit thermocouple calculation.
Because thermocouple do not fail regularly (i.e., the identity of the specific failed thermocouple does change on a frequent basis), taking points out of scan will not be a normal occurrence. Leaving the points in scan and causing an SPDS alarm condition will in effect provide an alarm condition that a thermocouple indication has failed and needs to be investigated.
"3.4.3 Security Only selected computer systems personnel have the passwords necessary to enter input or to make software change.
This provides a high level of system security. TVA should develop an administrative procedure to ensure that control room operators are notified when computer points are being manually entered or software modifications are in progress."
TVA Response i
Manually entered data shows on the SPDS display screen as a manual data point. Software changes are controlled by Sequoyah Standard Practice SQE-13
" Software Configuration Control," and permission is obtained from the shift operations supervisor before the system is removed from service for software modifications.
q I
l V
c l',
l "3.4.4 Operational Availability Curing the aud.'t, the following computer points were noted to be reading erroneously:
Containment Temperature RHR Containment Spray Valve position Containment Pressure Radiation Levels Feedwater Pressure 2 of 4 AFW Flows
=
All MSL Flows The radiation computer point problems are apparently being caused by some extremely long signal cables to the SPDS which are ungrounded and suffering degradation through current drain to ground.
TVA instrumentation and control personnel understand the fix for this problem (installing isolation transformers and grounding of the E-MAX isolation devices).
TVA should expedite repairing the remaining radiation instrument inputs to the SPDS/0TSC."
TVA Response TVA reviewed the status.of the modification that installed the isolation devices in the radiation monitoring points.
This modification is field complete.
"3.6 Designed incorporating accepted human factors engineering principles 1.
There are two separate concepts included in the SPDS design. CSF trees which begin to function 60 seconds after a reactor trip, are primary displays for emergency operations. The bar charts and alphanumeric displays are used to support normal operations.
Currently the SPDS is not designed to provide an overview of 5 required CSFs during normal operations since the software for the CSF trees is not operational during normal operations.
(NUREG-0700 Guidelines 6.5.1.1.a and 6.5.1.1.b)
The OTHER display on the RAD sections of 2PS1 Plant Status display is not descriptive enough to understand the data points during the alert.
(NUREG-0700 Guideline 6.5.1.1.b)"
TVA Response As noted in the response to item 3.1, the SPDS design concept has been changed so that the SPDS status trees are valid both pretrip and posttrip and the status of the SPDS is presented on every screen.
The bar charts are still available to the operator while in the SPDS mode as can be seen on the SPDS Map, figure 9, but are not used to determine the CSF status.
L
[.( :.
The point labeled OTHER in the. radiation section of 2PS1 has been relabeled EFFLUENT, and the monitors that feed this point have been changed.to the various plant airborne effluent radiation monitors.
"3.6 Designed incorporating accepted human factors engineering principles 2.
1FRZ F-05 limits are different on the SPDS.and hard copy, i.e.,
95% sump level on SPDS vs. 60% on hard copy. and 1000 R/hr containment on SPDS and 10 R/hr on hard copy. A software change-request is in place to modify the limits on SPDS.
(NUREG-0700 Guideline.6.5.1.1.a)"
.TVA Response As noted in the response to item 3.1 and as can be noted on figure 5, this has been corrected.
"3.6 Designed incorporating accepte1 human factors engineering principles 3.
Displayed action values on F-04 are different from the hard copy, i.e., 2400F on SPDS vs. 2450F on hard copy and 2700F on SPDS vs. 2750F on hard copy. A software change request is in place'to modify software to correctly display the same updated limits.
'(NUREG-0700 Guideline 6.5.1.1.a)"
TVA Response As noted in the response to item 3.1 and as can be noted on figure 4, this has been corrected.
"3.6 Designed incorporating accepted human factors engineering principles 4.
Last terminus box on SPDS CSF is labeled CSF SAT but the hard copy is labeled OK.
These are inconsistent.
(NUREG-0700 Guideline 6.5.1.4.a)"
TVA Response As noted in the response to item 3.1 and as can be noted on figures 1 through 8, this has been corrected.
"3.6 Designed incorporating accepted human factors engineering principles 1
l l
5.
The bar charts limits are set different than the CSF limits, i.e.,
1HCN Heatup/Cooldown indicates red containment Temp-Press while the CSF box indicates green normal.
(NUREG-0700 Guideline 6.5.1.6.d)"
l TVA Response The different limits for the containment pressure-temperature on the bar charts Lnd the SPDS containment status tree are intentional. The limits on the status tree are geared toward ensuring the integrity of the containment as a fission product barrier by ensuring pressure remains within design limits.
The precsure and temperature limits on the bar chart display (1HCN and 12R0) alarm blocks are designed to aid the operator in non-LOCA conditions where he is concerned with not entering an LCO.
Either abnormal temperatures or pressures will cause the combined pressure-temperature block on the bar charts to indicate an alarm condition.
"3.6 Designed incorporating accepted human factors engineering principles 6.
Numerous radiation process monitor points in the SPDS use units of Deca-Kilo Counts per Minute (DKCPM) which is very. unusual and different from the units of CPM used on control room instruments.
(NUREG-0100 Guideline 6.5.1.2.b)"
TVA Respcase The units for the referenced process radiation monitors are deca-counts per minute, not deca-kilo counts per minute. Although this is not the normal unit used by operators, it is not difficult to multiply the readings indicated on the technical support center data system (TSCDS) screen by 10 to get the normal units used by operators.
TVA agrees that it would be better to have the units consistent with control board indicators.
In the initial response to the audit, TVA indicated plans to make a change to the TSCDS data base to correct this problem by October 31, 1988. However, in the process of attempting to make the change, TVA discovered that the problem was more difficult to correct than anticipated.
TVA subsequently committed to resolving the problem with inconsistent units on the computer and control board instruments as part of the.'otalled control room design review (DCRDR) corrective action.
TVA still plans to make this correction and plans to have it complete by the end of the cycle 4 refueling j
outages, but is committed to having it complete by the end of the cycle 6 l
refueling outages.
l l
l l
\\-
"3.6 Designed incorporating accepted human factors engineering principles 3
i 7.
Magenta dashed. lines on CSF screens are Orange on the hard copy CSF pages.- The operators are trained to respond to a flashing terminus box rather than color alone. The reason for using magenta is because of the limited number (8) of colors available for Aydin display generators.
(NUREG-0700 Guideline 6.5.1.4.e)"
(
TVA Response SQN Operations has determined it is better to have the minor color mismatch than to change the hard copy pages and retrain operators.
i "3.6 Designed incorporating accepted human factors engineering principles 8.
Screens in the control room are not labeled 1 and 2.
Therefore the operator does not know which screen he is selecting on the control panel.
(NUREG-0700 Guideline 6.6.1.1)"
TVA Response Labels will be added to the CRTs as part of the corrective actions to the DCRDR. This is scheduled to be complete by the end of the Unit 2 cvele 4 refueling outage.
"3.7 Minimum information displayed should be sufficient to determine the plant status with respect to five functions Collectively, the six CSFs and Radiation Alarms described above are missing as inputs parameters listed below.
Inclusion of these parameters provide a comprehensive set to monitor the five Safety Functions listed in NUREG-0737, Supplement 1:
Main. Steam Line Radiation Reactor Vessel Level Indication Containment Hydrogen Concentration Containment Isolation Status" I
TVA Response As noted in the response to item 3.1, containment hydrogen has been added to the SPDS, and reactor vessel level will be added by the end of the unit 2 cycle 3 refueling outage. Main steam line radiation monitoring will be added when these radiation monitors are installed during the cycle 4 refueling
outages. As noted in the NRC audit report, containment isolation status is not required on the SPDS because the indications are already available in the control room in a well-grouped display in a location convenient to plant operators.
"3.7 Minimum information displayed should be sufficient to determine the plant status with respect to five functions The title 'OTHER' used on one of the radiation alarms is not descriptive and, in fact, the operators interviewed did not know what the alarm uses as inputs. Although the three radiation alarms available in SPDS/0TSC contain a comprehensive set of inputs (except Main Steam Line Radiation and Containment Radiation), TVA should consider relabeling and reorganizing the inputs for use in a top-level display. A frequently used scheme is to create a single radiation alarm which contains inputs from all of the major plant effluent points which have potential for being a source of a significant release to the environment. At Sequoyah, the list of inputs would include:
Air Ejector Exhausts Containment Purge Exhaust Waste Gas Decay Tank Effluent Monitor Auxiliary Building Vent Monitors (only Iodine channel appears to be used now)
Steam Generator Blowdown Radiation i
Main Steam Line Radiation The remaining building area and atmosphere monitors could be used in an alarm (s) which is anticipatory to an actual release to the environment."
TVA Response l
TVA has relabeled this point as EFFLUENT as noted in item 3.6.
The points monitored are:
Condenser vacuum exhaust (equivalent to air ejector exhausts suggested by NRC)
Shield building vent (containment purge and waste gas decay tank exhaust)
Auxiliary building vent Steam generator blowdown Service building exhaust TVA will add main steam line radiation monitors to this point when the monitors are installed as part of the PAM upgrade scheduled during the cycle 4 refueling outages.
"3.7 Minimum information displayed should be sufficient to determine the plant status with respect to five functions The Sequoyah SPDS uses a computed average (or bulk) cold leg temperature in the pressurized thermal shock (PTS) CSF algorithm.
IVA should evaluate the validity of using an average value'versus an auctioneered coldest or individual loop Tc in this algorithm (i.e., to monitor the worst cace nozzle or vessel downcomer beltline area)."
TVA Response The SQN SPDS uses auctioneered algorithms for the PTS status tree. For the entry level block addressing cooldown, the SQN SPDS cold leg temperature algorithm takes the highest cold leg temperature and the lowest cold leg temperature observed in the last hour and determines if the difference is greater than 100 degrees F.
These do not have to be in the same loop.
For all other blocks on this status tree, SQN SPDS uses the coldest cold leg temperature to determine necessary actions.
"3.7 Minimum information displayed should be sufficient to determine the plant status with respect to five functions The use of one -5 to +60 PSIG containment pressure instrument in the containment pressure-temperature alarm algorithm causes spurious alarms since the setpoint is around 2 PSIG.
TVA should consider using a narrow range instrument for input-to this algorithm."
TVA Response TVA has always used the narrow-range instrument for this algorithm. The narrow-range containment pressure as noted on the SQN SPDS agrees with actual containment pressures.
"3.7 Minimum information displayed should be sufficient to determine the plant status with respect to five functions If TVA makes the modification to have the CSF algorithms run continuously, reactor trip status (e.g., Reactor Protection System relays or Reactor Trip breaker position) will need to be added to the SUBCRITICALITY CSF."
j TVA Response As noted in the response to item 3.1, this has been incorporated.
"3.8 Procedures and operator training addressing actions with and without SPDS The review team evaluated the training module for SPDS and found some sections missing or in error.
For example, the radiation alarm 'OTHER' is not described in the training and the descriptions of other radiation alarms are not accurate as to the algorithm inputs. An SPDS very i
similar to the one installed at Sequoyah is also installed at the simulator where TVA personnel receive their training.
Two Senior Reactor Operators (SR0s) and one STA were interviewed during the audit.
During a typical SPDS audit, at least six SRO/STA personnel are interviewed. Because of limited time and the unavailability of SR0s/ STAS, conclusions on the level of training are based on a limited sample.
The following SPDS training deficiencies were noted:
SR0s/ STAS were unaware of differences between the CSFs and the FRGs.
SR0s/ STAS did not know that algorithms for CSFs are not running until after a reactor trip.
SRos/ STAS did not understand the inputs to the three radiation alarms.
SR0s/ STAS did not know that the superscript
'M' on a value means that the value has been manually entered.
SR0s/ STAS thought that Magenta backlighting of a parameter means it is suspect data when it actually means that the value has exceeded an alarm setpoint."
TVA Response As stated in TVA's previous response, training modules have been revised and incorporated into operator training. Operator training on SPDS was completed during week 2 of 1988 operator requalification.
When major changes are made to SPDS, operators receive retraining on the changes. Additionally, operators utilize SPDS during simulator training.
l f
i
,,u,
,.,w.
w w vu nv.
-- u a v -..cz-.
.* ~
~
-'~
==r*iu M ~m W T 4
o.
N5 s9h;Nk h d
~ ~
i
.u
- St ci Figure'l i
1 1
1 k
I I
{
.i EF 08 I
N N
Q*
Q "Y
05 Od Su 8"
ed od a
os v
v OL
'36 O
06 O
O e
..g i
E ti y
}
I l
1 M
O W
C W
1
- 2 2
i I
ss a
s.w s.
eCa tw c.
o I
Nower -
o.
i aw a-zaogs ao es J3 l
EHL '
m.W
'3
/'
3 W
g b
E' b:'
84 s3.8 H
we C
L QM e
W M 1.
g 4
=W
.)
Uwc wa3Oq r
002 m et up F-Z Q w 3Ew w
ICww OGI Maw Qa MQNr i.
t 1
-6
-~
1 C
g g g wo
~
Wh35 ikE(WI($W:
n s
v.
y gg
..e......
8 o
e 1
4 e
e 6
l g gg M V Z Q N M W Q j w.
h
+4-*
e-
,e g..e-
...w.
w,.
,.. :t 4 Y A\\)U&lM '
9*
s, f,
f k'
fSi/i;
's
(
1l1jj yk u_
t;.i yfrQyfip
. uith I
s
~~
%=
~
,t
=
N P" g af N D
r
.. v
< Y n&yN
.w%.'
8 5
1 1
2 O.
C C
TC h
GR f
GF i
h.
B i
i E
Is E
t E
E E
E a
e m
m m
m i
ru d
e Yf u
M l
E
}
b-
{
[
'l T N S
I A D
XC E/g PS T
t S pxSM Oe Ctq(
.s' I
G i
f I
GNIE
.r LT OA g Ou T
CQ qL L
sBE CUD aSA p
II g.
T N g.
h D
I 4F XCg E/T
,ii,; :5 N
E
.B.;.Mr MI". yjn.
Il
,ij!<
T 0 E S0 3:iI i
(i g.
RxS2 pj.
O%E1 i.
~
l
~
3 C P 2 E o. 3H
[
r.
CtL l
da t1
/ b < 3 + y O p'. [ $4a$<i;,hu4 j:'
3p% D d~ sN e.
4 t n,'
t 2
)
- 99u r x
3$w;n jc g*
sI~
a c ;
aS e
c.
M.dpi #1,Tk'.,A34.N g$%';Tk py.E
- ?-
g A,
s1 m.
g[d1 3
.}
i, f
.,4 4
i.
s s
- .. 4 h. [.
r.
~
l' f-3
$".?
i
- h:~
- ';g
.p' ~
u3, W @ a y.?
b T.
L'
. /
.[
7
~*
_4
,f[ll 1il l
C.s*
'g i
j
I w
me M,.s,,ygggyg.0y.ry g. l j
4
-u.qc:
i 4
4 8
l
,g
,.g (Mh.
- w" s v I
e Figure 3 1
i l
4 i
\\
r?,
P h
N T
Q
- l-b Y
Y j
Cm OG C3 03 04 X
06 d6 ou Ow Qu Q
j a :_.
M 7
./r
.lg
=
u a
4 at q
'I
.i i
'I E a s
I E'
M
~
8%$u; MS*
g g
w w
?
6, a s 4
azed
~
p5)
. - ~
s g
=
4
.OtAul J
W % Wf*
Q J~
w
' J7 2 W gjaa Wr?t 1.
. -u 3
T ud a
>g E dd${f ; @g 0 s
, ess-oa a
z h.zI a 1
O E
h:.
,......,.......,..............y..
g g
g$qqw. !.............I.....!...4........
h ut G
e e
e a
a e
a e
-i===
L" GWW%Qm l @,
E J.Juld M V Z 4 N M L e.I O
ae,
..,(,.g.-,,,.
-g/.,,..,,. -yw. 3 q* <
, g,,., 1,;(>Q (fj..r,
'*L-a
.{, &&'
r*
(,,
M,),.f,3,
^'
~
p.
.g J.
.m,.
r,=,_
s
.i}s
,. f.
- t' y; -
4
- .p I
4.
csg' g.,g' e,.g.;s.
y
.s,y.
4' Figure 4 f
l
- %bsthkm M;.'
M
+
If
- F o,
4 x
s x
g q
y w 4
Sh Sk Sh hk hk
' 'Y 82 85 SE 8
S$ 92 M
M 2
l
.e 3
n J
s t a t 1
.I l
I g
a i
a e e
s-4~
48 m
m
~
8-g N
g
'5 j i
E
~
~
L>
n 1'
. lo:ig s
r 5
- 7., y e.,
Di em m
$ 55 E
~
(/
-~
.g;_p-e n.,
gg g
g asy m
.#ph J>*.
a
- urL,
=
a< -oer o
t! NEdWR h505ES W5 n.
7
,.,4
~_
W %# 2 C N M L4.J Q
_m
'TT (QY' t ' v ** *K*y ***fWWhy m 3
- * * * " m"~ 'v3 v*;
qK
$4+ yt zNr u?.
- W
- s p
+ ~,
- n lt k1s f(Lk}J.
~'f
, ei( '
}
...r.
_ 7;s, r,
.%M,,y
@ @f; $ $ [ ff
$h(fjgMA y
-~,
g;-
3-y,yg ep il.'
s e/,j %A'
. sc J r, yu..
>.n.--
.g
)
8 i
Figure 5 i
I
)
b
.3 I
u a
n N.
hN.
kN.
hP!
h a,.
f, 85 8"
8" 85 880 M 5
l l
[!
l 98
=,_ w I
I i
e r
l Q
4 I-W
'v e z
zum O>W
.:p.
U$dG wo E E E!6
[
I
+
go a
ma-y Ew MJ I
l (AM Ss
$ 2'a8
+
n E
"l s.
n=
("
- a a a 5
5 *a ut"u" B'
4 s
is e
v (w]a ed g
-s-m m gg m
!7*,$,,.......,............,.......,!,5, !
i i
5a Q
vas.
m v 2:
a.
N M L.s c v_
g, 3
1
.. 7 7,.g.
.f N.7 r
.u w p ig[ # '~~
_. b;;'Y..,Q c,
N.., 1."
p 3 - Nik. k, g, Q; f.. f Qj;f l
- r. ? f ', e
- ~>*
. nec..
- c.. e :
,y.
,c
+a ya 9
Y'
~. fog -
f
- j,( '
^
.e.
m w. n m m :. g,c..;,. a
- ,,, sn~
7
,4*
.., e n.d :'g;gj~/.
J,{g f;,
[h
.d,.s,pg f 4
e 4
e a,
e Figure 6 l
l
(
1>
[* '
l 1;-
1,
. ",hp
ly'..
,, yp.
$.Y.
~
5
$N.
heb m
"?
OG 05 Om W
W
-J 04 t.N 04 O_
O
.;S E
88
[
u b
$l il C
g; I
il Z
g, li g
C N
II il Z
H M
.I GN M
We c.:
m WM JH 3 '
O W
)
NM MWG Z.
amj Z
e MWat WJW4 p.
E.Jof GW42 vir i 3
3 O CLJQH O.
W WJ M
N
@ Z
~ =
W a
$ 0, kh us c.
V
's w'
.Je m
e su m
n.
p d
rl--
Mac Cur WaM 5,,-j
.J Z
b b
~
>aa FA e.
mm hm 7.....,.......,......p......
=
5 WW wg 4
e e
i e
a e
e m v 2 c..
N M taJ c i e
'I
.mL
,n v-.
-,.,. < -.,.,.,-.:,,.,,,p7-,qr,-.. m m;.m prp; spy m 7-.7
?.
' \\
, &f.)f,e*}'f,-<j/p Y. L i..),& *.5' jc
=:*.iir*
'l ff$id@%,f $h.,, ((f'Y;'j?
-?'
.'fp,
g-
,.g;y
,.g.,
qtk
%1&f;g.
a
, +...
y-
4
'a f.
iL Figure 7 l
l A
.ij
$;y
~
F
~
~
~
~
~
89 S8 88 88 88 8
a es l,
N
- l
>E 8o-p,..
va5 o-WIZ 4
WW a-s M
WQ4 Q
ObO W Ok m
ZZ Ss Was
--s M
324 Q
Manw M'
Ef x-.
334 g
G eaW WJ IN
-W 0
ba5
!. U.,.,
M V Z b b M
O j'
}-
l l
l l
l Figure 8
)
i 1
1 I
l l
l W
4 r +.
. ~.
M j g 2NG J2M M
4 wam O
C
- i. l.
il
- i e
i
!i
~
a I f
i m
3 O
GN
^
ds 3
o cx w 30 if EG
<m=.mma-me ne e emmm mens.-.mm.,
-ew--
'---+eo
+p.-e,.
,eu a.
.gw-.w-
%a (UN wmma
.J U U Q 1
C q vv.J
/ 4 6V a
i e
e e
a e
e m v = A N ~ w o Li
. v-
=
T 7,1 EH c
GG AI C
PR
.G A
S P
T U N
U T
A e
A TK N
tT AN E
P EI St S
HS S
ER RO PF C
/
T G
N E I P L O O
[-j C
O 1
2 C
P P
E E
~
O O
K D
D N
3B b
0 S
T t
R T
S F
1 S
O I
E P
N l
T R
R R C
C P
I R
H A
R T
T E
U R
S
~
S Y
D G
T P.
2 N
AA S
/
S C
I T E
T E
1O A
D I
H M
ER R
D I
T E
N O
PL T
b4 P
O 9
N C
i T
t N
Tsti ED 6r I
A LA LR LD r
R
'l E
i P.
Pt t
Y M.
T U 0 R
O D
,t S
L T
I' 0
A D
O N
E P
O X
i H
t ll S
C ET N
Gr I
(
AE PL I
T
5 0
Figure 1 r ej N
N
.J.
O =
c.
a-O a a
Fw Fm Fu Fu I
a O m' e
e e
og og y
Og y
y OL Ou ou O
06 O
_O
{J u
il)
I m
m
'2'.:
('~5 (4)
~)
gg
?
LT.
r*
I N
a E
a H
OD W 11 a
I WB J Q
@&y
-1 L-H W G =4 QW H N uw H La
@WGZ I a Q G G' 000G C' '
- ) -
a r cg o
.: y F--
ZGQWII OGWW w ow MGLZH
@krar d
v1 r
[J e
m a
m
- e. n L.m b
I7
(.d
( 2J
- k..a. J
(%5 M
Z; Z'
W E1 S!
Yr 4
O g
H aa c3 t; -
a i5 m
W*8-W -
- m. + vWo a
14 aW -
[i 3$
$$N 20 E2HW wac 6-sii
_?
M CL Q G G W~
t M
"hW W ()
Z C1.
N W
Le.1 C
Figure 2 LI W
e N
e.
o.
o.
?A S
S S
02 03 OG V
- J' 06 06 06 O
M~f
. ~,
t' J)
J I
1 L'
ill ill
?
III 4
4:
Q~
\\
de-W
!ll ta u,
t--
C pJ Z
H T Lf1 m a O
XUg*6 En W L/I G 4XGQ 03 WO-V L.J
'?
- =
,.u>
o i
k W
JF o:
OG L;'h' ;
mmW Q3 F'--
va UDQ GWG t.,, ;
hg h
! Q C
H
.,..... S..........., {
[,-
'"'~ !
[
.f..
g W"w8
! E.....R...., E....!E !.,
!E E.bEi!
gymm s
a
...a a.
u.
Q. We e
e e
i e
e a
e J
m o x a.
m - w c=,
j cv m
o Figure 3 J
1 0"'
o".
C '.
O S.T
- z
- T rf rf
/3
'1 C3 03 O@
0%
CU M
fj CW 04 Ok 06 06
_O e,
s) z LA ms z CJ Q 55 E az a q
3
~-
?, _
8%L BA"#
i;-
z~O iG W
GM 3.-j[)
e j --
c; 4
m _
W OWQ WJ arsw-fh CL M M J az a
83 2 E
E @
4 as=
e u s wr-l 0
ZJG4 W
aw bJ h
' #j g
wn n
^-
m q. m,.,
U OBe 3
~
a r
T uO ua H W O
[h
~.'
3 z Gl
- OOw a,1 p.
r.
a OJs324 FLGOH yp g
CZ w- -
OG G
Lu.
z
- 1 O
g-O a
. [
r=.
8"O E, I k, k, k, N I.=..,..'i k, N.,..] N l!
,%Cw
.m.
awmsom j
zJJmo m u Z A N M
LI o m
S E
g
- I Figure 4 l
- l c
we N
hd hd hd h@i 85 8$
S$
s 85 85 s
s q
w w
m al at g
a j
..n u.
f[
L A.J i~~ ~)
h A.)
r i
j 5
V 11 ep dl wu a viu a "a
}
w w
u et sn auko avee g
l w
e, e-w u
.k h h O
=
8: D w
w w w3 au Ds
~s m
[E.
Pw ofew e
an.
%gr -
e o
a m
ww g "t ag
- duk3 a,
k%
<. w
[g 7 [3 e
o Od.ae%"W Ng w
m lt ao -
/
wgv 2a a v 2
aa y
y, s.
g d w r.E t i(#
,aa-wn woga-
- =
s-ao I:
$!dWR$
v5$3ES sP
- 2. 0 C'
[
A,.......,.,.,......,.,, E, I E. ! E i!
i ' 5, 1...,..!. -...........,
'1 CC m
m v 2 A N M 6.4J c
- g
i D'
.s.
Figure 5 I
1 m
ui m
n et, EN, EN.
EW RW E..
Ou Om 03 Om OQM M
fj 06 06 OL 06 OWce O
n
?
11 h
t 5
g
?
ll I:
D E
in ll p
H 1.
m 34 i
O W
I h
=
>~
U H
l Z
ESb
{$
H 3
q
- ggr, b,e ll t-FEM
~
g C
ZGAW
'u, a El ast; G
24 t
s
?
Vmmu ha b GY vara g
Ug
@wf X
a a
a-
?
I l
ma.
o o
rgu E2' 520.
g 4
umoO E
i T
g yi 64 NWh H
C r a s ta G.
H3Vk
& R 2
Gw w
h 5.w.
BQ
-2 a
t_~
,$ 4 04 A c.
r N
3
'f
YW
[
Ci ELS o-o gag 2
Ewd C
i i
j '- ' :
i i
QW,N L..
..;.....a d...i. i 4.
..il.... a G~
l C.A 4.J
- Z:
Q.
N M 64 o m
{t
t Figure 6 es Vi m
e N
O.
O+
O, e.-H HH
>H M.
e a
e OCg OCK O2 V
W c6 ou ou O
O O
v1 j '.
7 t,
T ll k.
U to p
p y
U k
7;)
!i
^-
11 g
7 li LT
)
N D
CD M
be
& E b-g-
t.n s 4W MWGZ M
3 W bi-l pt/l
@JH"
- -)
C W
(p. :
Z.
r MW st WRWG ow*
G.Jf U s1 3 CR GWQZ kJQH
- >O W
WJ W
N WWE H 2 WPG
-F G Q GWI wJ 3 H kJ&
@JJ W
N A
QWWQ i
4 JiMI du Ln 1:s; I b
C3
@@Q Wj GUI WaN g
f.
wh Lli
=:
r3
~
wj Goa
./-
LD u
Q d
ZW 1
O MV H
r 1' "--*r****1 Z
w UU gg s
e.
...a.....4
.e
/gM w
WW 5
1 0
t i
4 I
e
^*
On CJ Z Q N H LJ Q.
--cn
.-.,.---.-------.r,-.-----.---
8 g
s Figure 7 i
1 (J
e 9
m' 7
C0 00 00 oc oc y
v-I
~7
~
e L
t, n
d p
n 1
i F
9 i
Un at p
15 P
C p'
6 4
JC 4
k b
g, y
q at GH I
C
'9 Oh g
U h
m x
me s
O W
inn U
p W"k QH I
a-1:
WZZ c~.
-+
WW 4
n
.-4 M
W34 11 Q
U44W Qb KE f.M I. ;.-.;
C:n il Om LV
.JQ VJ4 Mmq 85 k
334 Q
LAmW cs Se k3 -
+
as XH4 a
L Z
w WJ I$
e
-u 4.J g
CJ 5x to o
u
._.J l
LA CG Q J
7.-..7...........).......,..
7
- i.. -...i....i. -...i.....!.5..'....!...... -....... ]
u
.m 1
e i
e e
e e
4 s
m u Z A N M L.6J c
w I
l
e
_,, _, _ _ - - - - = - - - -
--,-m------
g
'd h
'Ab
=
. 3 Q..
Figure 8 cn~
.m 6,
%a Saw x
vam a
g N
Y.
r!
k II i
N f
U h
h 0
il 2a h
?"I I
a-t-
l Ld M w
U IB f4 tr YM '
cn Sd a
A se LA La s
a e
(E W 30 (E 4 U,
E C.A t-ge G.
La J fum mmma C
wWWW t>
3 sV QJ O
Q-C T
W N
G
==
ll1:ll 11 3 :,
>q >
MO g T
7 2
- 31 EH 2
AI
!E PR kGA S
F T
I U $s A' U N
T-A TK h TN An LTn r
ANk E
P EI [
St S
~
HS $
/
I T
j GN EI(
R CL O
E O
, :+
C Ol C
1 2
Pv-PD.
DBr
- E E
O' V
O.
D M
O 4
OB KL H
S SO P
t R
T S
?'
EM f' 5
O I
R TP I.
R R
I Y
I h k
C C
PH R
- f A
C E
U i T
T R
S SD G
YT [!
P 2
M A Ah S
/
a TT 9I_
E S
C I
EI 1
c -
A O-I H'
D M ~3. <
ER, R
T I 3L 7"
D E
N T k O
PL O
,F+
t I
t i
C I'
T N
N TNO l
}
E F
UDk LD M A
A L
I R*
F PA F
R E
f P
Pf m
U Y
L O,,
R TD e,,
6 O
}
S L
T P
D A
D O.
N 2 E
P O
E f
H S
C U
I H
E ET M
1 GF I
(1 1
AE l1 1
PL 9
. s-Il flllllIlll u