Links to articles:
CDC - Technical Considerations when Bringing Hemodialysis Facilities Water System Back on
Line after Hurricane Katrina
CDC - Remediation and Infection Control Considerations for Reopening Healthcare Facilities
Closed due to Extensive Water and Wind Damage
CDC - Healthcare water system repair following disruption of water supply
CDC - Key Facts about flood recovery
CDC - Safe Use of Tanker Water for Dialysis
CDC - Guidelines for Environmental Infection Control in Health-Care Facilities (see page 59)
CDC - What to do when your municipal water supplier issues a boil water advisory - dialysis clinics
EPA - National Primary and Secondary Drinking Water Standards
Northwest Renal Network - Monitoring your dialysis water treatment system
Water Conductivity vs TDS - Chart
Water Conductivity vs TDS - Table

Below is a list of water treatment issues seen by Florian Services at
Hemodialysis Clinics:

Facility
Backflow Preventer not re-certified yearly per code.
Water heater not inspected yearly.
No remote water alarm in patient treatment area (DI Low Resistivity alarm, Low RO Tank
alarm, RO alarm, etc).
Remote alarm found with the speaker muted.
Windows in the Water Room causing bacteria to grow in the water system.  Found green
growth inside polypropylene RO Tank and flow meters (clear rotometers).  Windows were
covered with a light blocking film when the water system was installed, but still saw green
growth.
Windows in the Reuse Room causing degradation of the Renalin®.
Water heater is set above 140F.
Safety showers are not tested monthly.

Mixing Valve
Check valve leaking.  Hot water flowing into the cold water lines at the mixing valve.
No bypass valve.

Booster Pumps
Pumps are too small; can not support RO and a regenerating softener (Duplex softener).
2 pumps with no check valves after each pump causing water from the online pump to back
feed through the off-line pump.  This caused loss of water flow and pressure to the RO and
lead to failure of the pumps due to the offline pump turning backwards (motor failure due to
too high amps needed to get the pump turning the correct direction when the pump is
started).
Bladder Tank after the Booster Pump is too small causing the Booster Pump to cycle
frequently.  This could lead to premature failure of the pump.
Bladder Tank failure causing the Booster Pumps to cycle frequently.  This could lead to
premature failure of the pump.
Pressure switch not set correctly.  Causes Booster Pump to cycle frequently while RO is
running.

Pretreatment (general)
Times for backwash/regeneration are not staggered causing some to occur at the same
time.  If softener bypasses during regeneration, this can leave hard water in the carbon
tanks and deliver hard water to the RO the first time it runs.  If the softener does not bypass,
this will prevent the carbon tanks from being backwashed.
No lockout on RO during backwash/regeneration
pH meter out of calibration.  Check pH meter at least weakly (if pH control installed).
Using clear-braided hose.  No clear hoses should be used.
No UPS for filter heads.  A UPS will prevent the clocks from being off if power is lost.  If the
clocks lose power, the filters can regenerate/backwash during dialysis hours.  If power lost
during backwash, clocks will cycle through backwash, but Booster Pump will not run.  May
effect Softener regeneration.
Regeneration/backwash orifice found defective allowing too much flow.  This could possibly
backwash out some of the filter media and dilute the salt used to regenerate the softener.
With the RO off, the pressure gauges on the pretreatment do not read similar values.  This
indicates that one or more pressure gauges are out of calibration.

Sediment Filter
To high of a pressure drop (>10 psid) with no action.  Check pressure gauges or increase
backwash frequency.
Filter contained in a clear housing.  This can promote bacteria growth.  All filter housings
should be opaque.

Softener
No salt in the Brine Tank.
Not sized to allow degradation over time causing measurable hardness after several years
of operation.
Pressure drop too high (>10 psid).  Check pressure gauges or increase backwash time
during regeneration.  Check maximum flow specification for the softener to ensure it is not
exceeded.  If no sediment filter upstream, install sediment filter.  If sediment filter is
installed, rebed to decrease micron rating of filter.
Iron in raw water reduces the life of the resin.  Need to rebed more frequently and/or use salt
with iron-out.  The raw water analysis should always include iron.
Hardness after the softener not monitored at the end of the day (single softener).
For softeners that regenerate based on throughput (duplex softeners), hardness after the
softener not monitored at the end of the allotted throughput.  Need to record the gallon
throughput value when taking the hardness sample.
Hardness sample should be taken from the RO reject water instead of downstream of the
softener.
No lockout on RO when regenerating.  Lockout is required.
Small duplex softener regenerates twice per day (each tank once per day).  There may not
be enough time for the salt to dissolve for the next regeneration (requires at least 4 hours to
dissolve).
High sodium in the raw water (300 ppm) decreasing the efficiency of the softener to remove
calcium and magnesium.
Using Potassium Chloride salt to regenerate softener instead of sodium chloride salt.  
Using KCl could adversely affect the patients.  AAMI guidelines need to be revised to specify
the use of Sodium Chloride salt and specify not to use Potassium Chloride salt.  Home
dialysis patients especially need to be training on this requirement since Potassium
Chloride salt is available in many stores.

Carbon
Carbon Tanks only achieve 6 minute EBCT (or less).
RO requiring more flow than it was designed for (someone increased the reject flow).  This
is reducing the EBCT in the carbon to below the 10 minutes required by AAMI.
Piping for replaceable carbon tanks not configured in a first in/last out configuration causing
too high flow through one train (won’t meet 10 minute EBCT in this train).
Three trains of carbon tanks.  One train has a different size tank (smaller).  Will not get even
flow through each train and will not get 10 minute EBCT on train with small tank.
To long of a time between backwashes (should be at least every 3 days).
Bypass valves installed.  Bypass valves on carbon tanks serve no purpose except to allow a
chance to leak potentially causing a positive chlorine result.  Remove bypass valves.
Carbon is not acid-washed.
No certification of the iodine number for the carbon installed (need Certificate of
Conformance for that lot of carbon).
No lockout on RO when backwashing and the carbon will bypass when backwashing.
The length of the piping for carbon tanks piped in parallel is not the same length.  This can
lead to uneven flow through each train of carbon filters potentially causing the EBCT to be
reduced to below the AAMI guideline.
Unit using the wrong size chlorine packet to perform the chlorine test (HACH kit – using 5ml
packet for 10ml sample).  This will result in an improper analysis possibly resulting in a
missed positive result.

RO
Brass fittings on RO potentially leaching into the product water.  Minncare® will corrode
brass.  Brass fittings were on feed water side (Continental RO).
Chip in the Osmonics OSMO 23G RO is not Y2K (need version 6.9 or higher).  Upgraded
chip will also allow periodic flushing of the RO while the RO tank is full if the RO has a
permeate divert capability.
Improper alarm settings (set maximum conductivity, minimum % rejection, maximum
temperature, maximum pressure, etc. to the company’s standard).
RO sized too small (not sized for full operation of all dialysis stations or not sized for
expansion).  Increasing the size of the RO may impact the amount of carbon required to
achieve a 10-minute EBCT.  Softener may also be affected if more water is processed.
Has string wound prefilter.  String wound filters should be avoided due to chemicals
(surfactant) used on string.  RO Tank found full of foam (soap).
No permeate divert system on RO.
The RO is cleaned with an acid and caustic cleaner, but not disinfected (need to use
disinfectant such as Minncare®).  Acid and caustic cleaning does not disinfect.
The RO has not been cleaned in over 1 year.  RO cleaning should be part of Preventative
Maintenance (at least every 6 months).
Concentrate piping routed below the RO membranes allowing the membranes to drain
when the RO is off.  This creates an air pocket in the membrane and could cause a
pressure spike when the RO pump starts possibly rupturing the RO membrane.

RO Water Storage Tank
No filter on the vent.  The tank should have a 0.2 micron filter.
Loop-seal in overflow pipe.  Bacteria can grow in this loop-seal.
Water is fed to the RO Water Storage Tank when bypassing the RO.  This will contaminate
the RO Water Storage Tank with bacteria from the Carbon Filters.  To bypass the RO, water
from the Carbon Filters should be fed directly to the DI Tanks (DI Tanks must be followed by
Ultrafilter).  Isolate the end of the loop to prevent the water pressure from opening the back-
pressure valve and filling the RO Water Storage Tank.
No spraying device inside the tank (from RO or from the Loop return).  A spraying device is
required to prevent bacteria from forming on the tank walls and to help rinse out any
disinfectant.
Insufficient flow coming back to the RO tank or the tank is too large in diameter for the spray
ball to spray the water onto the sides of the tank.

Loop Distribution Pumps
Don’t have check valves after each pump, which allows some water to back-flow through the
pump that is not in use.  This caused a loss of flow through the loop and can lead to
premature failure of the pumps due to them spinning backwards when offline (too much
current required to start pump due to it spinning backwards).
Pumps made of cast iron, which can leach into the dialysis water.  Pumps need to be made
of Stainless Steel.
Pump too small to create a turbulent flow (need at least 3 feet/second flow).
Too large causing too much flow through hollow-fiber filters or too high pressure (DI Tanks
only rated to 100 psi).  Can correct by installing a recirculation loop to the RO Tank with a
back-pressure relief valve.
Only 1 pump installed with no backup plan.  Failure of pump can impact dialysis operation.
Pumps do not trip on low level or low flow.  This can damage the pump.
Pumps are swapped too infrequently.  This can lead to bacteria growth.  The pumps should
be swapped daily.

DI tanks
Conductivity/resistivity meter not reading correctly. (Reading over 18 Megohm-cm)
No analog/digital resistivity meter after the DI tanks; only has a 1 Mohm-cm light. DI Tanks
release fluoride when exhausting.
If installed, the DI tanks typically are not required to process 100% of loop flow (split flow is
acceptable and will reduce costs).  If a split flow system is installed, a flowmeter should be
installed to monitor the flow through the DI tanks.  
DI tanks left in-service too long.  If DI tanks are installed, they can grow bacteria
(recommend maximum life of 3 months).  
DI Tanks followed by 0.2 micron filter or 5 micron filter.  DI Tanks must always be followed by
an Ultrafilter (0.05 micron or less).  If ultrafilter is a spiral-wound membrane, should install a
5 micron filter between DI Tank and Ultrafilter to catch resin so the resin does not plug the
inlet of the spiral-wound membranes.
RO bypass line is hard piped with no way to tell if the isolation valves are leaking (hard
piping can be installed with 2 isolation valves with a drain between them).  Hard-piped RO
bypass should also be short since the water in the line will be stagnant for long periods of
time allowing bacteria to grow.  If the RO bypass is not hard-piped (bypass line is a hose
installed at the time of the bypass), the hose should be disinfected before use.

UV Light
UV installed; recommend removing.  UV does NOT remove bacteria.
No ultrafilter after UV; only 0.2 micron filter.  An ultrafilter (0.05 micron or less) is required.
UV lamp not changed periodically (yearly).  
UV lamp not disposed of properly.  UV lamp contains mercury.

Ultrafilter
Pressure gauges measure pressure drop across filter AND the flow restricter.  Flow
restricter will add to pressure drop.  Should only measure pressure drop across filter.
Not piped in a first in/last out configuration causing too low of a flow through loop piping.
Wrong valve turned to flush spiral-wound UF causing flow to the loop to be temporarily lost.
0.2 micron filter installed.  This does not meet AAMI.
No final filter installed.  Ultrafilter required if have RO Water Storage Tank, or DI, or UV
installed.  Ultrafilter can be installed on dialysis machine.
Not compatible with some disinfectant chemicals.  Bleach used to disinfect loop.  The
bleach went through a spiral-wound ultrafilter that is not compatible with bleach.  The
integrity of the ultrafilter may have been jeopardized.
Unit is told the final filter is 0.05 micron by water vendor, but is actually 0.1 micron.
Not enough filters installed to provide the proper flow through the loop to support dialysis,
reuse, and have enough to create a turbulent flow to prevent biofilm formation.  Need at least
3 feet/second flow.
Spiral-wound filters flushed during treatment hours.  It will reduce the pressure to the
dialysis machines.

Loop Piping
Not configured in a loop
Dead-legs greater than 1.5 times pipe Inside Diameter.
Bladder Tank installed on a direct feed system (no loop).  Bacteria can grow in the Bladder
Tank.
Loop configured so the Reuse Room is fed first and then the dialysis area.  The Reuse
Room should be fed last.
Pressure at the end of the loop is over 50 psi (Renatron® only rated for 55 psi).
Clear tubing used in the loop.  Clear tubing can expose the water to sunlight increasing the
possibility of bacteria growth.
Insufficient sample ports; can’t sample the beginning of loop, individual dialysis stations,
end of loop, bicarbonate tank fill connection, dry acid tank fill connection, etc.
Performing bacteria sampling using ¾” or 1” valves.  Can not adequately flush these valves
before taking bacteria sample.  Numerous false bacteria results observed.
Loop piping installed by plumbers not familiar with dialysis requirements: used 90-degree
elbows instead of two 45’s, not inserting the pipe all the way into the socket, cut pipes with
hacksaw rather than cutters and left burrs, etc.
Loop piping diameter increases instead of decreasing at the end of the loop.  Piping went
from 1.25” schedule 40 to 0.5 inch schedule 80, and back to 1.25” schedule 40.  The loop
had been modified several times.
Loop too long causing too high of a pressure drop and too low flow.  This could lead to the
formation of biofilm.  If have more than 32 stations, two loops should be installed.
Not being disinfected routinely.  AAMI requires at least monthly.  CDC recommends weekly.

Daily Log Sheet
Not being performed correctly (writing down the wrong information, writing down useless
information).  The pressure on the inlet of a filter was being recorded and the pressure on
the outlet of the filter was being recorded with limits for each, but the differential pressure
was not being recorded.
Not being completely filled out.
No corrective actions taken for parameters that are out of specification.
Chlorine analysis being performed wrong.  Even if chlorine were present, the solution would
not have turned pink.

Acid Feed System
Head tank too high (maximum 14 feet above dialysis floor) causing too high acid pressure
on dialysis machines.
Acid tank or header tank overflows due to acid crystals on the level switch.  The unit should
not have bleach or major components (pumps) under the header tank in case if overflow.
Acid being used is not for the same proportion as the bicarbonate.
Acid tank has an open lid.  This can allow things to fall into the tank (aluminum wrench, bag
of bicarbonate/salt).  A unit would fill the acid tank by removing the lid and placing a hose
into the acid tank.  To prevent the hose from falling out, a bag of salt or a bag of bicarbonate
was placed over the hose on the rim of the acid tank.  One day the bag fell into the acid tank.  
All of the acid had to be drained.

Bicarbonate Feed System
Not rinsed/disinfected daily.  (AAMI requires weekly disinfection)
Using raw water to rinse out bicarb jugs.
Bicarb conductivity and pH in a large bicarb tank are not checked at the beginning and end of
the day to ensure some of the bicarb is not lost during the day.
Bicarbonate being used is not for the same proportion as the acid.

Other
Iron and zinc are not being checked for in the raw water sample.
TOC and Silica are not being checked for in the raw water sample.  TOC can foul carbon.  
Silica can scale RO membranes.
AAMI RO water sample not taken when the RO has high conductivity.
RO Conductivity not recorded when the AAMI RO water sample is taken.
AAMI RO water sample taken soon after the RO has started.  Need to allow RO to rinse
down; take sample after RO permeate divert is completed.
AAMI water sample taken after the DI tanks instead of after the RO.
Bleach is being stored near Vinegar or Renalin® (acids).  Bleach and acids should not be
stored together.
Chrome-plated quick-disconnect being used to connect dialysis machine to the water
supply.  Clinic thought they were stainless steel.
Problems Identified in Water Treatment Systems at Hemodialysis Clinics
Florian Services, LLC
Protecting you against the hazards of water
Florian Services, LLC                                                    Phone: (630) 310-2510
www.florianservices.com                                                        Fax: (630) 910-0877