tech talk

Mixing Water Control

Part V of Concrete Quality Series
By Karthik Obla, Ph.D., P.E., VP, Technical Services,
and Colin L. Lobo, Sr. VP, Engineering

arts I and II of the Concrete Quality series1,2 discussed

P
measure the quantity of wash water retained in the drum. Therefore
that a good measure and benchmark of concrete quality is in compliance with ASTM C94 truck operators should be educated
the Standard Deviation (SD) of compressive strength test of the importance of disposing of all of the wash water prior to batch-
results. The primary factors that impact the SD are variabil- ing a fresh concrete load. The process of discharging the mixer wash
ity associated with materials, production and testing. In order to water into a wash out pit may take time but it is an important step
reduce the strength standard deviation the concrete producer needs toward achieving quality concrete. The more common industry prac-
to manage those aspects of variability that can be controlled. Part tice is to not wash out the mixer drum after every load. The mor-
IV3 discussed and quantified the concrete strength variability due tar coating will not significantly change the water content or any of
to variation in water and air content. It concluded that in order to the properties of the new load of concrete. Washing the inside of a
attain a low SD of compressive strength, among other things, it is mixer drum may be necessary in hot weather conditions when it
important for the producer to attain a low variation in water con- can lead to undesirable hardened concrete buildup inside the drum,
tent and air content as they contribute significantly to the varia- with special concrete containing color or fibers, or when required
tion in strength. This article discusses how a producer can manage by company policy.
the various factors that will reduce the variability of mixing water
content in concrete mixtures. Batch water
Batch water is the water added to the ready mixed concrete truck.
Mixing Water Generally this constitutes the primary source of mixing water. Water
Compared to other ingredients, there are several sources (and can be measured by volume using water meters or volumetric tanks; or
potential errors) by which mixing water is incorporated in a concrete by weight in scales. ASTM C94 requires that the added water should
mixture. ASTM C944 and ACI 1175 state that the total mixing water be accurate to within ±1% of the design total mixing water. The
should be within ±3% of the quantity required by the mix design. NRMCA plant certification process requires that volumetric water
ASTM C94 states that mixing water includes water added to the measuring devices should be accurate to within ±1.5% of the desired
batch, ice added to the batch, water occurring as surface moisture on amount of water which corresponds approximately to an accuracy
the aggregates and water introduced in the form of admixtures. The of ±1% based on total mixing water for typical aggregate moisture
primary sources of mixing water variation and methods by which it levels. Scale accuracy requirements are tighter. The NRMCA mea-
can be controlled to be within a tight tolerance are discussed below: suring tolerance is more easily verified during a batching process or
by review of batch records. The NRMCA plant certification process
Washwater in truck mixer drum from previous load requires that batching measuring devices be verified for accuracy at
When concrete is discharged at the jobsite a mortar coating stated frequencies or when there is reason to assume their accuracy is
adheres to the surface of the mixer drum. Depending on the situa- in question. It is one thing to have a well calibrated water batching
tion, type of mixture or company practice, the concrete mixer truck device. It is additionally important to ensure that the added water in
operator may wash the interior of the drum at the jobsite or on arrival every concrete load that is shipped from the plant is within tolerance.
at the plant. In some cases, this water may not be discharged and Batching errors do occur from time to time. All computerized batch
remain in the truck when the next load is batched. Washing the mixer panels have some form of error monitoring and alerting system which
drums between every load leads to several possible problems: may be overridden and accepted with a keystroke. Automated evalu-
1. It consumes a lot of water; ation and alerting systems available nowadays can send e-mail alerts
2. If discharged, it generates process water at the concrete plant to Quality Control/Quality Assurance (QC/QA) personnel when
that will have to managed in compliance with environmental a load is found to be outside prescribed or company-set tolerances.
regulations; Improving batching accuracy has been discussed at a great length
3. If not discharged, water (anywhere from 5 to 30 gallons) may elsewhere.6
remain in the mixer drum when the next load is batched. Most chemical admixtures added to concrete contain in excess of
ASTM C94 recognizes the last possibility and states a condition 60% water. It is important to consider the water content contributed
that the wash water retained in the drum for use in the next batch by chemical admixtures, particularly those which are used in high
should be accurately measured. However, it is not easy to accurately dosages such as corrosion inhibitors or accelerators. In general, if the

CONCRETE in focus ı 23
admixture dosage is around 1 gallon per cubic yard or greater, batch on water hoses; using high pressure, low volume spray bars; installing
water compensation should be made to account for water incorpo- water meters on the water lines or using water from the truck water
rated with admixtures. For the admixtures that are used in high dos- tank for retempering to obtain a measurement of the added water. The
ages, if the dosages change significantly between loads and the water goal should be to minimize the water added to the mixture at this point
from the admixtures had not been considered then it may lead to a and to be able to record the quantity of water added to the mixture. It is
high mixing water variation. important to educate the truck operators to minimize the water enter-
ing the mixer drum and record the amount of water added.
Free Water from Aggregates Automated water addition devices installed in truck mixers are
Free water from aggregates constitutes a portion of the mixing also available. These devices introduce recordable quantities of water
water. If the coarse aggregate and the fine aggregate have a free mois- to the concrete mixture based on calibrated slump meters that measure
ture content of 1% and 5%, respectively, assuming a typical concrete the mixture consistency. These devices can be controlled to shut off
mixtures that contains 1800 lbs of coarse aggregate and 1200 lbs of when the mixing water reaches the quantity permitted for the mix-
fine aggregate, the contribution that the aggregates provide to the ture. Automated admixture addition devices are also being explored.
mixing water is 78 lb/yd3. If the total mixing water is 270 lb/yd3 the
free water from the aggregates contributes 29% of the mixing water in Water added at the job site7
this example. Therefore, an error in estimating the aggregate moisture When the truck arrives at the job site and the concrete is about
content can cause a significant variation in the mixing water content. to be unloaded it is quite possible the concrete slump is lower than
To address this many concrete producers use moisture probes in the that specified or desired by the contractor for placement. Th is
fine aggregate bins. These moisture probes provide real time mois- depends on the time elapsed from batching, ambient and concrete
ture measurements of the fine aggregate based on which the added temperature, and mixture proportions used, among other factors.
water is adjusted automatically or manually. NRMCA plant certifica- ASTM C94 allows a one-time addition of water so that concrete
tion requires that moisture probes be verified for accuracy at defined slump can be brought to specified levels provided the design mix-
frequencies or when there is an assumption of error. Alternatively, ing water content is not exceeded. It is common practice for pro-
aggregate moisture should be determined at the concrete plant using ducers to “hold back” water to permit job site water addition. Th is
a hot plate or a microwave oven according to ASTM C566. If such a water addition should be done before any significant quantity of
process is followed the NRMCA plant inspector guide suggests that concrete is discharged; C94 allows for slump and air tests on a
the aggregate moisture be measured not less than 3 times per week. preliminary sample to facilitate this. C94 requires that the water
This can vary depending on aggregate storage methods or frequency added by the receiver of concrete and his initials be recorded on the
of receiving shipments. Moisture probes are not commonly used in delivery ticket. At the request of the purchaser, C94 requires that
coarse aggregate bins due to wear. Coarse aggregate moistures should the delivery ticket has all the information required for calculating
be periodically measured using a microwave oven or a hot plate. Since the total mixing water. It is a good practice for producers to note
they constitute the largest quantity of ingredient batched, a small the amount of water that can be added at the job site so that design
change in the coarse aggregate moisture can result in a large change water content is not exceeded. Adding about 2 gallons per yard over
in the mixing water content. Another important factor is the proce- the design amount can result in a loss of the strength overdesign for
dures used by the loader operator to feed aggregates from stockpiles the specified strength. ACI and ASTM standards do not establish
to the batch plant. The key is to avoid loading aggregates with variable accuracy requirements for water added from truck water tanks.
moisture and to ensure that the aggregate being batched is represen- The NRMCA Plant Certification has accuracy requirements for
tative of the moisture content assumed in the batching process. In measurement devices on truck water tanks. Sight gauges can gen-
general, working the complete face of a stockpile tends to average out erally measure accurately to ±1 gallon and water meters on truck
the moisture content and minimize variation. water systems can do better.

Water added at the “slump rack” Variation in Mixing Water Demand Due
Almost all concrete is accepted at the job site based on a certain to Temperature, Delivery Time and
slump level or visual consistency to facilitate ease of placement, Air Content Variation
regardless of whether the project specification includes a slump As mentioned earlier, concrete at the job site is expected to be
requirement. After loading all of the materials, truck mixer operators within a specific slump range before it can be placed. Temperature,
move the truck to the wash down area or slump rack to clean the load- delivery time, air content and a host of other factors affect the mix-
ing hopper, chute, and sides of the mixers before leaving the plant. ing water demand of a given concrete mixture that needs to meet a
At this point they visually estimate the slump and might retemper specific slump level.
the load to achieve the desired mixture consistency, including some Table 1 summarizes the eff ect of temperature and delivery
anticipated slump loss during transit. In some cases actual slump time on mixing water demand from a previous NRMCA research
measurements may be made and/or slump meters on truck mixers study 8. The study shows that if a mixture had been designed at a
may be used. But typically this adjustment is done on a visual basis concrete temperature of 65°F and a delivery time for 20 minutes
that is subjective to the truck operator. Water added at the slump an increase in temperature to 95°F and delivery time to 90 min-
rack can be a large source of variation of mixing water in concrete utes can cause an increase in average mixing water content of 33
and needs to be managed by proper company policy and education lb/yd 3. Th is variation in mixing water can clearly lead to increased
of truck operators. Some options to control this include using timers SD in strengths and low strength problems. The concrete producer

24 ı MARCH/APRIL 2011
has limited control over concrete temperature and very little con- Water Variation and Slump Variation
trol over the delivery time. For a given concrete mixture if 10 different truck loads have slump
The producer can reduce the mixing water variation due to tem- varying over a tight range (4 in. to 5 in. for example) would it be
perature variation through the following steps: 1. Table 1 shows that reasonable to expect that the mixing water content is controlled to a
for every 10°F increase in concrete temperature the water demand tight range? The answer to this question is “yes,” provided everything
increases by 5 lbs/yd 3 (about 2% of the mixing water amount used else stays the same.
typically) assuming a delivery time of 60 minutes. So the producer Consider the job site data for the following 2 truck loads that used
could increase the dosage of water-reducing admixtures as the con- identical materials and mixture proportions:
crete temperature increases by more than 10°F. 2. Concrete mix- Load A, Concrete temperature=65°F, Air content=7.5%, Delivery
tures that are designed at a concrete temperature of 65°F cannot time=20 minutes, Slump=4 in.
be used when the concrete or ambient temperature exceeds 85°F. Load B, Concrete temperature=65°F, Air content=4.5%, Delivery
When the temperature reaches and exceeds 85°F adjustments, such time=90 minutes, Slump=4 in.
as the use of retarders and increased amounts of pozzolans like Class From the discussions in the previous section it can be quickly esti-
F fly ash or slag cement, allow for longer slump retention without mated that Load B is likely to have about 35 lb/yd3 higher mixing
excessive water addition. water content than Load A even though the slumps measured at the
Variation in mixing water demand due to delivery time varia- job site are the same. If the temperature of Load B is 20°F higher than
tion is much harder to control since the producer cannot predict that of Load A it can be estimated that Load B is likely to have nearly
the delivery time of any particular load in advance. Based on Table 45 lb/yd3 higher mixing water content than Load A! In addition other
1 concrete producers should aim to hold back about 2 gallons/yd 3 variables such as presence of fines in aggregate, dust on the aggregate
(or target a slump of 2 in. higher than the minimum permitted surface, variations in cementitious shipments can also influence the
at the job site) to compensate for slump loss that can occur due water demand, thus resulting in loads that are similar in slump but
to the delivery time. There will be no variation in mixing water that differ in mixing water content. With this kind of potential dif-
demand due to delivery time variation if the concrete mixtures do ference in mixing water demand for two loads with the same slump
not exhibit any slump loss. As explained earlier, use of retarders and one wonders about the basis of rejecting concrete at the job site on the
increased amounts of pozzolans like Class F fly ash or slag cement basis of slump.
allows for longer slump retention, but one has to be careful adopt- Slump measurement at the plant (not at the job site) does have an
ing those techniques when concrete and ambient temperatures are interesting benefit for the producer. Let us say a producer targets a
below 85°F due to potentially delayed setting times. Job site admix- slump range at the plant which is the same as at the jobsite and holds
ture (water reducer) addition may also help reduce the variation back 15 lbs/yd 3 to compensate for the slump loss that can occur due
in mixing water demand due to delivery time variation provided to delivery time. If the slump measured at the concrete plant prior
qualified personnel are available to administer that. Automated to any water addition at the slump rack is out of that target slump
admixture (water reducer) addition devices can also be considered range for that mixture it is a clear indication that either the mix-
when available. ing water content is not within tolerance or the water demand for
ASTM C94 and ACI 117 state that for air entrained concrete the that mixture has changed due to a change in characteristics of the
tolerance is ±1.5%. Figure 1 extracted from the NRMCA Concrete materials, batched quantities, concrete temperature or air content.
Technologist Certification program shows the typical reductions in When a slump measurement is out of the target range it is a good
mixing water content expected due to the entrained air content. idea to conduct a quick aggregate moisture test either through the
Added air illustrated in the figures is the difference between the microwave oven or a hot plate or using a Speedy moisture meter.
target air content in air entrained concrete and that present as If the aggregate moistures are as designed then it is clear that the
entrapped air in non-air entrained concrete with the same materi- problem lies elsewhere:
als. Air entrainment causes greater water reductions for low cement 1. Was there wash water in the mixer drum?
factor or higher w/cm mixtures. If two successive concrete loads 2. Was the batch water out of tolerance? (Typically, batch water is
have a total air content at the concrete plant of 7.5% and 4.5% (just 200 lb/yd3, so the maximum possible variation in mixing water is
meeting the tolerance of 6±1.5%) the added air for each load can be ±3 lbs/yd3 if batch water is in tolerance)
calculated as 6.0% and 3.0% and therefore the difference in mixing 3. Was the air content out of tolerance? (Typically, the air content
water content can be estimated from Figure 1b as 20 lbs/yd 3, assum- tolerance is ±1.5%; this can lead to a variation in mixing water of
ing a design w/cm of 0.50 for that concrete mixture. There is a silver ±10 lbs/yd3)
lining in this though. Generally, a higher air content results in a 4. Was the concrete temperature more than 20°F different from the
lower mixing water demand, so the resulting strength reduction due design concrete temperature for that mixture? (From Table 1 it
to the higher air content is somewhat reduced. Another consequence can be estimated that for every 10°F increase in concrete tempera-
of this is that when water is added and mixed into a relatively fluid ture the water demand should increase by 2 lbs/yd3, provided the
air-entrained concrete batch at a moderate temperature, additional slump is measured at the plant within 10 minutes after batching
entrained air is generated thereby increasing the slump more than the materials)
anticipated by normal rules of thumb. Once again job site admix- If the answer to these questions is “no” then changes in material
ture (water reducer) addition or the use of automated admixture characteristics, such as excessive fines in aggregate, dust on the aggre-
(water reducer) addition devices can help reduce the variation in gate surface, variations in cementitious shipments or batched quanti-
mixing water demand due to air content variations. ties, should be evaluated. If air contents and concrete temperatures

CONCRETE in focus ı 25
Table 1. Average Mixing Water Increase Due to Temperature and Delivery Time (from Reference 8)
Condition Average Mixing Water Increase, lb/yd3
Delivery time maintained at 20 min. and concrete temperature increased from 65°F to 95°F 12
Delivery time maintained at 90 min. and concrete temperature increased from 65°F to 95°F 19
Temperature maintained at 65°F and delivery time increased from 20 to 90 min 14
Temperature maintained at 95°F and delivery time increased from 20 to 90 min 21
Temperature increased from 65°F to 95°F and delivery time increased from 20 to 90 min 33

Mixing water increase was averaged over 8 different mixture types (2 different cements, 2 different strength levels, with and without fly ash). In all
cases mixing water content was adjusted to attain a slump level of 4±1 in. at discharge.

Table 2. Sources of Mixing Water Variation and Means to Reduce it

Sources of Mixing Water Variation Recommended Practice
Washwater in truck mixer drum from previous load Avoid washing the mixer drum after every load.
If that is not always feasible educate truck mixer operators to discharge all of the wash
water prior to batching a fresh concrete load
Batch water out of tolerance Calibrate water measurement devices to be accurate to ±1.5% of the target quantity and
verify accuracy at least every 90 days and more often if necessary.
Educate batchmen not to override batching error alerts.
If possible invest in an automated alerting system for out-of-batch tolerance that can send
e-mail alerts to QC personnel to address a load before it reaches the customer.
Aggregate moisture Have moisture probes in fine aggregate bins if possible and verify for accuracy at least every
90 days.
Do not ignore measuring coarse aggregate moisture content.
If probes are not used, use a hot plate or a microwave oven and measure aggregate moisture
and update batch computers at least three times a week.
Adjust added water in all loads to account for changes in aggregate moisture content.
Stockpile aggregates and load aggregates into the plant to minimize variations.
All of the above are an NRMCA plant certification requirement.
Water added at the slump rack Establish company policy and equipment for slump rack water adjustment. Consider
adding the water from truck water tank or install water meters in water lines. Record the
amount of water added.
Water added at the job site ASTM C94 allows a one-time addition of water to enable slump adjustments.
Note the amount of water that can be added at the job site so that design water content is
not exceeded.
Truck operators should be required to record job site water addition and obtain the
signature of the purchaser or representative on the delivery ticket. Ensure measuring
devices on truck water tanks comply with the NRMCA Plant Certification accuracy
requirements.
Temperature, Delivery Time, Air Content and Other Increase the dosage of water-reducing admixtures to target a 2% mixing water reduction
Factors for every 10°F increase in the concrete temperature increases by more than from the design
temperature.
Concrete mixtures that are designed at a concrete temperature of 65°F should appropriately
be modified when the concrete or ambient temperature reaches or exceeds 85°F.
Concrete producers should aim to hold back about 15 lb/yd3 (or target a slump of 2 in. higher
than specified) to compensate for slump loss that can occur due to the delivery time.
Consider mixtures with good slump retention performance without delayed setting times
to address delivery time variations.
Consider job site admixture (water reducer) addition with qualified personnel at the job
site to address delivery time and air content variations.
On a daily basis QA/QC personnel can utilize the slump measurements at the plant prior
to any water addition after batching as a quality assurance tool to ensure that the mixing
water content is within tolerance or ingredient characteristics are unchanged. Alternatively,
daily records of water added to the mixture at the slump rack can be used.

26 ı MARCH/APRIL 2011
100 100
Added Air Added Air 8% 6%
8% 4%
80 80

6%
60 60
4%
2%
40 40
2%

20 20

0 0
300 400 500 600 700 800 900 0.40 0.50 0.60 0.70 0.80 0.90
Cement
Cement Factor, lb/cu.yd
Factor, lb/cu. yd Water-cement Ratio
(a) (b)
Figure 1. Effect of added Air on Mixing Water Demand (from Reference 9) (a) when cement factor is known;
(b) when w/cm is known

are not measured one could still do the above exercise but the inves- temperature can be nearly 45 lbs/yd3! So steps as discussed in this
tigation process should be started when the measured slump varies article are needed to reduce this variation in mixing water demand.
more than 1 in. from the target slump range. The 1 in. allowance is By following the various steps, a producer focused on quality should
mainly to account for the air content and temperature variations. be able to maintain a good control on the mixing water content and
It is impractical to measure slump at the plant prior to water addi- meet the ASTM C94 and ACI 117 tolerance of ±3% of the quan-
tion for every load that leaves the plant. It may be sufficient to do tity required by the mixture proportions. On a daily basis QA/
that every 10 loads. For the remaining loads visual slump estimations QC personnel can utilize the slump measurements at the plant
can be made by the truck operators. These slump estimations and prior to water addition at the slump rack as a quality assurance
measurements should be captured and analyzed on a daily basis by tool to ensure that the mixing water content is within tolerance.
QC/QA personnel and any large deviations from target slumps can Alternatively, daily records of water added to the mixture at the
be investigated as discussed above. Instead of measuring slumps at the slump rack can be used. ■
slump rack one could also look at the daily record of the water added
to the mixture at the slump rack. If more than 1 gallon/yd3 is being References
added it can mean one or more of the following: 1
Obla, K.H., “How Good is your QC – Part I of Concrete Quality Series,” Concrete
InFocus, May-June 2010, Vol. 9, No. 3, NRMCA, pp. 17-18.
1. Mixing water content is not within tolerance 2
Obla, K.H., “Sources of Concrete Strength Variation – Part II of Concrete Quality
2. Water demand for that mixture has changed due to a change in Series,” Concrete InFocus, July-August 2010, Vol. 9, No. 4, NRMCA, pp. 21-23.
3
Obla, K.H., “Variation in Concrete Strength Due to Water and Air Content Variation
characteristics of the materials or batched quantities – Part IV of Concrete Quality Series,” Concrete InFocus, January-February 2010, Vol.
3. Air content is below the tolerance level 10, No. 1, NRMCA, pp.
4
ASTM C94/C94M, “Standard Specification for Ready Mixed Concrete,” Volume
4. Concrete temperature is more than 20°F above the design con-
4.02, ASTM International, West Conshohocken, PA, 2010, www.astm.org.
crete temperature for that mixture 5
ACI Committee 117, “Specifications for Tolerances for Concrete Construction and
5. The target slump range is again being exceeded; Materials and Commentary,” American Concrete Institute, Farmington Hills, MI,
2010, 76 pp.
investigations as outlined before can be carried out. 6
Bain, D., and Obla, K.H., “Concrete Quality Control: The Untapped Profit Center,”
Concrete InFocus, Fall 2007, Vol. 6, No. 3, NRMCA, pp. 63-69.
7
CIP 26, “Jobsite Addition of Water,” Concrete in Practice series, National Ready
Summary Mixed Concrete Association, Silver Spring, MD.
8
The different sources of mixing water variation and the means to Gaynor, R.D., Meininger, R.C., and Khan, T.S., “Effect of Temperature and Delivery
Time on Concrete Proportions,” Temperature Effects on Concrete, ASTM STP 858,
control it is summarized in Table 2. The producer can employ the T.R. Naik Ed., American Society of Testing Materials, West Conshohocken, PA, www.
suggested practices to carefully calculate the mixing water content astm.org.
9
NRMCA Concrete Technologist Certification Instruction Manual, “Outline and
in the concrete. However, the variation in mixing water demand Tables for Proportioning Normal Weight Concrete - Advanced Version,” NRMCA,
due to variation in concrete delivery time, air content and concrete Silver Spring, MD, 2010.

CONCRETE in focus ı 27